Neural solvers for Vehicle Routing Problems (VRPs) have shown great advantages in solving various kinds of problem types. However, they also face critical challenges in generalizing from small-scale training to large-scale problems and in identifying the most salient topological information for decision-making. To mitigate these gaps, we introduce ScaleNet, a novel hierarchical framework that integrates a U-Net architecture into a unified, multi-task VRP solver. Scale-Net explicitly captures multi-scale structural patterns by processing a nested hierarchy of input graph instances. This enriched, coarse-to-fine representation is extracted by the encoder and fed directly into the decoder, empowering decoder module with superior topological awareness for routing decisions while simultaneously reducing computational overhead in the encoder. We conducted extensive experiments on 16 VRP variants with instances ranging from 50 to 5,000 nodes. The experimental results show that Scale-Net demonstrates significant performance gains over state-of-the-art baselines across in-distribution, zero-shot, and real-world settings.

3D Gaussian Splatting (3DGS) has demonstrated impressive Novel View Synthesis (NVS) results in a real-time rendering manner. During training, it relies heavily on the average magnitude of view-space positional gradients to grow Gaussians to reduce rendering loss. However, this average operation smooths the positional gradients from different viewpoints and rendering errors from different pixels, hindering the growth and optimization of many defective Gaussians. This leads to strong spurious artifacts in some areas. To address this problem, we propose Hard Gaussian Splatting, dubbed HGS, which considers multi-view significant positional gradients and rendering errors to grow hard Gaussians that fill the gaps of classical Gaussian Splatting on 3D scenes, thus achieving superior NVS results. In detail, we present positional gradient driven HGS, which leverages multi-view significant positional gradients to uncover hard Gaussians. Moreover, we propose rendering error guided HGS, which identifies noticeable pixel rendering errors and potentially over-large Gaussians to jointly mine hard Gaussians. By growing and optimizing these hard Gaussians, our method helps to resolve blurring and needle-like artifacts. Experiments on various datasets demonstrate that our method achieves state-of-the-art rendering quality while maintaining real-time efficiency, yielding LPIPS improvements of 5.1%, 19.7% and 6.3% on Mip-NeRF360, Tanks&Temples and Deep Blending, respectively.

In this paper, we aim to create physical digital twins of deformable objects under interaction. Existing methods focus more on the physical learning of current state modeling, but generalize worse to future prediction. This is because existing methods ignore the intrinsic physical properties of deformable objects, resulting in the limited physical learning in the current state modeling. To address this, we present NeuSpring, a neural spring field for the reconstruction and simulation of deformable objects from videos. Built upon spring-mass models for realistic physical simulation, our method consists of two major innovations: 1) a piecewise topology solution that efficiently models multi-region spring connection topologies using zero-order optimization, which considers the material heterogeneity of real-world objects. 2) a neural spring field that represents spring physical properties across different frames using a canonical coordinate-based neural network, which effectively leverages the spatial associativity of springs for physical learning. Experiments on real-world datasets demonstrate that our NeuSping achieves superior reconstruction and simulation performance for current state modeling and future prediction, with Chamfer distance improved by 20% and 25%, respectively.

The advent of Large Language Models (LLMs) has opened new frontiers in automated algorithm design, giving rise to numerous powerful methods. However, these approaches retain critical limitations: they require extensive evaluation of the target problem to guide the search process, making them impractical for real-world optimization tasks, where each evaluation consumes substantial computational resources. This research proposes an innovative and efficient framework that decouples algorithm discovery from high-cost evaluation. Our core innovation lies in combining a Genetic Programming (GP) function generator with an LLM-driven evolutionary algorithm designer. The evolutionary direction of the GP-based function generator is guided by the similarity between the landscape characteristics of generated proxy functions and those of real-world problems, ensuring that algorithms discovered via proxy functions exhibit comparable performance on real-world problems. Our method enables deep exploration of the algorithmic space before final validation while avoiding costly real-world evaluations. We validate the framework's efficacy across multiple real-world problems, demonstrating its ability to discover high-performance algorithms while substantially reducing expensive evaluations. This approach shows a path to apply LLM-based automated algorithm design to computationally intensive real-world optimization challenges.

Spatio-Temporal Video Grounding (STVG) aims to retrieve the spatio-temporal tube of a target object or person in a video given a text query. Most existing approaches perform frame-wise spatial localization within a predicted temporal span, resulting in redundant computation, heavy supervision requirements, and limited generalization. Weakly-supervised variants mitigate annotation costs but remain constrained by the dataset-level train-and-fit paradigm with an inferior performance. To address these challenges, we propose the Agentic Spatio-Temporal Grounder (ASTG) framework for the task of STVG in an open-world and zero-shot setting. Specifically, two specialized agents SRA (Spatial Reasoning Agent) and TRA (Temporal Reasoning Agent) constructed leveraging modern Multi-modal Large Language Models (MLLMs) work collaboratively to retrieve the target tube in an autonomous and self-guided manner. Following a propose-and-evaluation paradigm, ASTG duly decouples spatio-temporal reasoning and automates the tube extraction, verification and temporal localization processes. With a dedicated visual memory and dialogue context, ASTG achieves architectural efficiency by minimizing the number of reasoning calls compared to exhaustive per-frame reasoning, eliminating the logical redundancy inherent in joint-reasoning systems. Experiments on popular benchmarks demonstrate the superiority of the proposed approach where it outperforms existing weakly-supervised and zero-shot approaches by a margin and is comparable to some of the fully-supervised methods.

Large language models (LLMs) have shown promising potential for next Point-of-Interest (POI) recommendation. However, existing methods only perform direct zero-shot prompting, leading to ineffective extraction of user preferences, insufficient injection of collaborative signals, and a lack of user privacy protection. As such, we propose a novel Multitask Reflective Large Language Model for Privacy-preserving Next POI Recommendation (MRP-LLM), aiming to exploit LLMs for better next POI recommendation while preserving user privacy. Specifically, the Multitask Reflective Preference Extraction Module first utilizes LLMs to distill each user's fine-grained (i.e., categorical, temporal, and spatial) preferences into a knowledge base (KB). The Neighbor Preference Retrieval Module retrieves and summarizes the preferences of similar users from the KB to obtain collaborative signals. Subsequently, aggregating the user's preferences with those of similar users, the Multitask Next POI Recommendation Module generates the next POI recommendations via multitask prompting. Meanwhile, during data collection, a Privacy Transmission Module is specifically devised to preserve sensitive POI data. Extensive experiments on three real-world datasets demonstrate the efficacy of our proposed MRP-LLM in providing more accurate next POI recommendations with user privacy preserved.

MOTIVATION: High-fidelity simulation of nanopore sequencing signals is critical for rigorous benchmarking and validation of the nanopore signal processing pipeline. However, existing signal simulators often fail to capture the non-linear dynamics of nanopore current signals, relying on static pore models or lacking optimization objectives tied to basecalling, resulting in synthetic signals with low basecalling accuracy and fidelity. RESULTS: We introduce NanoSimFormer, an end-to-end Transformer-based signal simulator that integrates basecaller guidance during training to generate high-fidelity nanopore signals. NanoSimFormer achieves a median basecalling accuracy exceeding 99% and Q-scores above 22.8 for Oxford Nanopore Technologies' latest DNA R10.4.1 and direct RNA sequencing, closely mirroring real experimental baselines. It faithfully recapitulates experimental variant calling performance across the five human samples, achieving F1-scores of 0.9953-0.9973 and 0.7862-0.8612 for single-nucleotide polymorphisms and small indels detections, respectively. Compared with previous simulators, NanoSimFormer also substantially reduces false positives in homopolymer and short tandem repeat regions. NanoSimFormer-derived reads enable high-quality de novo bacterial assembly with consensus error rates below one mismatch per 100 kbp and maintain high correlations with experimental abundance in metagenomic and transcriptomic datasets. AVAILABILITY AND IMPLEMENTATION: NanoSimFormer is freely available on GitHub at: https://github.com/BioinfoSZU/NanoSimFormer.

This paper investigates the complexity of intelligent decision-making within multiagent systems operating under uncertainty, particularly emphasizing the challenges associated with modeling behaviors of other agents and optimizing decision-making for a subject agent in a common environment characterized by incomplete historical data. To address these challenges, we propose a generative learning method based on a general multiagent decision making framework, namely interactive dynamic influence diagrams, and apply contrastive learning to diversify the generation of potential behaviors, thereby enhancing the subject agent’s modeling and prediction capabilities. We conduct experiments on multiple classic domains to demonstrate the efficacy of the new learning method in improving decision-making quality. The empirical results highlight its substantial improvement in enhancing overall performance in multiagent decision-making. Our work contributes to multiagent decision making particularly when a subject agent interacts with other unknown agents, including humans, in many practical applications.

Adversarially robust knowledge distillation aims to compress a large-scale robust teacher model into a lightweight student counterpart while preserving adversarial robustness and natural performance. Previous methods primarily focused on aligning knowledge (e.g., predictions) between teacher and student models to transfer robustness. However, potentially incorrect predictions from the teacher can misguide the student, negatively impacting robustness transfer. To circumvent this, we propose a novel adversarially robust knowledge distillation scheme that promotes alignment towards more benign predictions rather than incorrect ones by refining inputs into so-called "inverse adversarial examples" via simply reversing the sign of adversarial perturbation. Through a comprehensive investigation of the properties of inverse adversaries, we provide new theoretical insights showing how mimicking the behavior of the teacher model on inverse adversaries facilitates reliable robustness transfer built upon the implicit connection between robustness and the input gradient information. We thus design a gradient matching mechanism between teacher and student models utilizing inverse adversaries to facilitate robust knowledge alignment. Furthermore, inspired by our analysis of the correlation between robustness and adversarial transferability, we propose a weight-space disruption strategy that jointly interacts with both teacher and student models to find a shared direction for better robustness transfer. Empirical evaluations across various datasets demonstrate that our method achieves state-of-the-art robustness and natural performance. Notably, on ImageNet, our approach outperforms prior methods by approximately 3.8% in both clean and robust accuracy. Moreover, we show that incorporating auxiliary generated data into distillation further boosts robustness. Our method can also be generalized to multimodal architectures.

The main focus of this work is on deriving a prescribed-time optimization control solution for a family of networked systems with uncertain nonlinearities and local time-varying cost functions. The underlying problem becomes much more challenging if the communication topology is not only local but also frequently switching. To overcome the technique difficulty arising from the local and switching network, a casted observer is introduced into the proposed distributed prescribed-time optimization control scheme, which allows the necessary non-local information to be localized within a prescribed time under the switching topology. Furthermore, the proposed prescribed-time optimization control algorithm is based on finite time-varying gain, avoiding excessive control input (especially excessive initial control input) caused by sustained high gain based method, without which the prescribed-time optimization result can not be derived. In addition, the gain switching time in the switching controller can be pre-assigned, distinguishing itself from most existing finite gain based prescribed-time control methods. Finally, the effectiveness of the method is validated through both numerical simulation and real-world experiment, demonstrating its potential for practical applications in networked control systems.

With the maturity of depth sensors in various 3D safety-critical applications, 3D point cloud models have been shown to be vulnerable to adversarial attacks. Almost all existing 3D attackers simply follow the white-box or black-box setting to iteratively update coordinate perturbations based on back-propagated or estimated gradients. However, these methods are hard to deploy in real-world scenarios (no model details are provided) as they severely rely on parameters or output logits of victim models. To this end, we propose point cloud attacks from a more practical setting,i.e., hard-label black-box attack, in which attackers can only access the prediction label of 3D input. We introduce a novel 3D attack method based on a new spectrum-aware decision boundary algorithm to generate high-quality adversarial samples. In particular, we first construct a class-aware model decision boundary, by developing a learnable spectrum-fusion strategy to adaptively fuse point clouds of different classes in the spectral domain, aiming to craft their intermediate samples without distorting the original geometry. Then, we devise an iterative coordinate-spectrum optimization method with curvature-aware boundary search to move the intermediate sample along the decision boundary for generating adversarial point clouds with trivial perturbations. Experiments demonstrate that our attack competitively outperforms existing white/black-box attackers in terms of attack performance and adversary quality.

Evolutionary algorithms (EAs) have gained prominence as a powerful optimization tool inspired by biological evolution, excelling in various complex domains. In the context of Generative Artificial Intelligence (GAI), EAs have shown promise in generating diverse, high-quality solutions. However, traditional EAs heavily rely on human-designed fitness functions, which may often lack flexibility and comprehensiveness for different GAI scenarios. Recently, the emergence of Large Language Models (LLMs) has opened new avenues for enhancing the evolutionary process. This paper proposes a formal framework named ECG-LFit, which utilizes an LLM (e.g., GPT-4-Turbo) for multimodal fitness evaluations. We validate the effectiveness of our framework using EAs (CMA-ES, MAP-Elites, and CMA-ME) in our case study on Super Mario game level generation. The results show that our framework improves the quality and playability of the generated levels. Additionally, user studies indicate that participants prefer the levels generated by the ECG-LFit framework, particularly regarding aesthetics, challenge, and playability. Furthermore, to enhance evaluation efficiency, we design a distilled model to simulate the scoring process of the LLM, enabling rapid and effective content evaluation in resource-constrained environments.

In the evolutionary computing community, the remarkable language-handling capabilities and reasoning power of large language models (LLMs) have significantly enhanced the functionality of evolutionary algorithms (EAs), enabling them to tackle optimization problems involving structured language or program code. Although this field is still in its early stages, its impressive potential has led to the development of various LLM-based EAs. To effectively evaluate the performance and practical applicability of these LLM-based EAs, benchmarks with real-world relevance are essential. In this paper, we focus on LLM-based recommender systems (RSs) and introduce a benchmark problem set, named RSBench, specifically designed to assess the performance of LLM-based EAs in recommendation prompt optimization. RSBench emphasizes session-based recommendations, aiming to discover a set of Pareto optimal prompts that guide the recommendation process, providing accurate, diverse, and fair recommendations. We develop three LLM-based EAs based on established EA frameworks and experimentally evaluate their performance using RSBench. Our study offers valuable insights into the application of EAs in LLM-based RSs. Additionally, we explore key components that may influence the overall performance of the RS, providing meaningful guidance for future research on the development of LLM-based EAs in RSs.

The multi-objective dynamic flexible job shop scheduling (MO-DFJSS) problem is crucial in modern manufacturing, impacting productivity and operational costs. Genetic Programming (GP) has emerged as a prominent method for MO-DFJSS due to its ability to evolve real-time responsible and effective scheduling heuristics. However, existing GP approaches often learn multiple heuristics for different regions of the Pareto front, making their management and selection complicated in real-world applications. This paper proposes a novel Pareto set learning GP (PSLGP) framework that addresses this limitation by learning a single, preference-conditioned heuristic that encompasses the entire Pareto front based on user preferences. This simplifies scheduling and allows for real-time adaptation to user-defined priorities. The framework employs a novel preference-conditioned heuristic representation that incorporates user preferences as additional inputs, enabling dynamic heuristic adjustments. To efficiently evaluate fitness without increasing training time, a surrogate model is used to estimate individual performance across different preferences, and three new fitness aggregation strategies are designed to ensure effective heuristic alignment across the Pareto front. Experimental results demonstrate that PSLGP significantly outperforms the state-of-the-art multi-objective GP approach, particularly in less busy MO-DFJSS environments, providing a more adaptable and efficient solution for dynamic scheduling challenges. Further analyses of preference influence, solution distribution, and heuristic structure provide evidence that the proposed PSLGP effectively learns preference-conditioned scheduling heuristics that align user preferences with various regions of the Pareto front.

Multiobjective optimization (MOO) is important for deep learning applications with multiple conflicting objectives. Pareto front learning (PFL) methods learn a single model conditioned on the preference of objectives and can be applied to any preference at inference time. However, existing PFL methods use predefined strategies (e.g., uniform sampling) to generate preferences, which could result in unevenly spaced solutions since the shape of Pareto front is largely ignored. In this article, we propose a lightweight and model-agnostic method to train a preference generator for a given PFL model, which learns to generate proper preferences from uniformly sampled ones, such that the resulting solutions are evenly spaced on the Pareto front. Compared to previous works, our method enables a more rational allocation of preferences, which can either be utilized to enhance a pretrained PFL model or be seamlessly integrated into the PFL training process to improve efficiency. We apply our method to state-of-the-art PFL methods with various backbones (e.g., multilayer perceptron, convolutional neural network, transformer) and validate the significance of preference generation across various tasks, from multitask supervised learning to multiobjective reinforcement learning-based neural combinatorial optimization. Experimental results show that our method improves the backbone algorithm in most settings, showing its effectiveness and general applicability.

Deep learning has been extensively explored to solve vehicle routing problems (VRPs), which yields a range of data-driven neural solvers with promising outcomes. However, most neural solvers are trained to tackle VRP instances in a relatively monotonous context, e.g., simplifying VRPs by using Euclidean distance between nodes and adhering to a single problem size, which harms their off-the-shelf application in different scenarios. To enhance their versatility, this paper presents a novel lifelong learning framework that incrementally trains a neural solver to manage VRPs in distinct contexts. Specifically, we propose a lifelong learner (LL), exploiting a Transformer network as the backbone, to solve a series of VRPs. The inter-context self-attention mechanism is proposed within LL to transfer the knowledge obtained from solving preceding VRPs into the succeeding ones. On top of that, we develop a dynamic context scheduler (DCS), employing the cross-context experience replay to further facilitate LL looking back on the attained policies of solving preceding VRPs. Extensive results on synthetic and benchmark instances (problem sizes up to 18k) show that our LL is capable of discovering effective policies for tackling generic VRPs in varying contexts, which outperforms other neural solvers and achieves the best performance for most VRPs.

Decentralized learning (DL) has proven to be effective for privacy-preserving next point-of-interest (POI) recommendation by sharing check-in information among users and collaboratively training on-device models. Existing works, however, simply assume that users tend to share check-ins with neighbors of short geographical distance or similar preferences yet ignore users’ actual willingness to share the information (WSI), causing potential privacy concerns. As such, we present a WSI-guided hierarchical DL framework for next POI recommendation (WHDL-Rec) to seek enhanced privacy protection with recommendation accuracy assured. In particular, WHDL-Rec first performs hierarchical data segregation to partition the private and public user data. It then accords to the server-client architecture, where the server exploits the public data to automatically learn users’ WSI w.r.t. check-ins and capture global user behavior patterns for recommendation accuracy maintenance; and the clients fuse the learned global patterns with the local private data for personalized on-device next POI recommendation, whereby WSI-guided collaborative learning is conducted with more secure check-in sharing. Extensive experiments on three real-world datasets demonstrate the efficacy of WHDL-Rec in delivering more accurate and privacy-preserved recommendations.

This work addresses the prescribed-time consensus tracking problem for multiagent systems (MASs) under denial-of-service (DoS) attacks, where certain communication links between agents are intermittently disrupted or rendered unavailable. Due to the DoS attack, although appropriate defense mechanisms may be employed to recover some of the attacked or backup connections, the communication topology evolves from a fixed graph to a time-varying switching topology. This ever-changing network introduces significant challenges, leading to discontinuities in the Laplacian matrix, which complicates both prescribed-time controller design and stability analysis. To tackle these challenges, a prescribed-time stability lemma (Lemma 5) is developed along with a vital inequality (Lemma 6) that establishes the quantitative relationship between Lyapunov functions across switching instants. Building on these results, a novel distributed observer is designed to accurately estimate the state of the leader within the prescribed time, despite the occurrence of DoS attacks. Subsequently, a coordinate transformation and a fractional-power backstepping technique are introduced to construct a control protocol that achieves prescribed-time consensus tracking. The proposed approach is rigorously supported by theoretical analysis and is further validated through its application to multiple unmanned vehicles.

Large language models (LLMs) provide a promising way for accurate session-based recommendation (SBR), but they demand substantial computational time and memory. Knowledge distillation (KD)-based methods can alleviate these issues by transferring the knowledge to a small student, which trains a student based on the predictions of a cumbersome teacher. However, these methods encounter difficulties for LLM-based KD in SBR. 1) It is expensive to make LLMs predict for all instances in KD. 2) LLMs may make ineffective predictions for some instances in KD, e.g., incorrect predictions for hard instances or similar predictions as existing recommenders for easy instances. In this paper, we propose an active LLM-based KD method in SBR, contributing to sustainable AI. To efficiently distill knowledge from LLMs with limited cost, we propose to extract a small proportion of instances predicted by LLMs. Meanwhile, for a more effective distillation, we propose an active learning strategy to extract instances that are as effective as possible for KD from a theoretical view. Specifically, we first formulate gains based on potential effects (e.g., effective, similar, and incorrect predictions by LLMs) and difficulties (e.g., easy or hard to fit) of instances for KD. Then, we propose to maximize the minimal gains of distillation to find the optimal selection policy for active learning, which can largely avoid extracting ineffective instances in KD. Experiments on real-world datasets show that our method significantly outperforms state-of-the-art methods for SBR.

Real-world multi-objective optimization problems often rely on physics-based simulators or physical experiments to assess solution quality, resulting in significant computational costs. In such scenarios, Gaussian process (GP) surrogate-assisted optimizers have demonstrated exceptional optimization performance. This article focuses on the theoretical convergence analysis of two existing decomposition-based GP-assisted optimizers: ParEGO with the upper confidence bound (ParEGO-UCB) and ExTrEMO. Unlike prior studies that typically assume a single weight vector for scalarization, this work primarily investigates multi-weight vector settings. Specifically, we analyze the regret bound of ParEGO-UCB within a rigorous theoretical framework and prove that, under a multi-weight vector setting, its convergence rate surpasses that of GP-UCB, which independently and sequentially optimizes multiple decomposed subproblems. Building on this foundation, we further explore the convergence properties of ExTrEMO, an expensive multi-objective optimizer designed for multi-source transfer optimization, in the context of multi-weight vector settings. Theoretical findings reveal that ExTrEMO achieves a tighter regret bound in multi-source settings compared to single-source scenarios, highlighting the advantages of leveraging additional sources to enhance optimization efficiency and convergence.

Human motion synthesis aims to generate plausible human motion sequences, which has raised widespread attention in computer animation. Recent score-based generative models (SGMs) have demonstrated impressive results on this task. However, their training process involves complex curvature trajectories, leading to unstable training process. In this paper, we propose a Deterministic-to-Stochastic Diverse Latent Feature Mapping (DSDFM) method for human motion synthesis. DSDFM consists of two stages. The first human motion reconstruction stage aims to learn the latent space distribution of human motions. The second diverse motion generation stage aims to build connections between the Gaussian distribution and the latent space distribution of human motions, thereby enhancing the diversity and accuracy of the generated human motions. This stage is achieved by the designed deterministic feature mapping procedure with DerODE and stochastic diverse output generation procedure with DivSDE. DSDFM is easy to train compared to previous SGMs-based methods and can enhance diversity without introducing additional training parameters. Through qualitative and quantitative experiments, DSDFM achieves state-of-the-art results surpassing the latest methods, validating its superiority in human motion synthesis.

Dataset condensation has emerged as a promising technique to improve data efficiency under limited data budgets.However, when applied to the text level, existing methods struggle to compress more information into samples through optimization.Thus, these methods provide no obvious advantage over simpler coreset selection despite their high computational cost.In this paper, we introduce CondenseLM, a novel paradigm for both effective and efficient textlevel dataset condensation.Our framework employs an LLMs-driven approach to sidestep the inherent limitations of existing methods, successfully generating more informative and less biased samples.In addition, it incorporates reward matching to align the LLMs-condensed dataset with the original dataset, maximizing representability and coverage.We conducted extensive experiments on SST-2, MNLI, AG News, and IMDB.Our approach outperforms both coreset selection and existing dataset condensation methods by large margins while also substantially reducing the computational cost.

Multitasking evolutionary algorithms can be effectively used to solve a number of problems with a single population. A key issue in deciding their effectiveness, is how to transfer good solutions from one problem instance to another problem instance which shares some characteristics. We investigate in this paper how to transfer solutions between different problem instances of the Travelling Salesperson Problem (TSP) based matching algorithms and introduce different transfer mechanisms based on matching the nodes between problem instances. In our experimental study, we examine how the different transfer approaches perform for different classes of TSP instances dependent on the characteristics of the considered problem instances.

Real-world optimization often demands diverse, high-quality solutions. Quality-Diversity (QD) optimization is a multifaceted approach in evolutionary algorithms that aims to generate a set of solutions that are both high-performing and diverse. QD algorithms have been successfully applied across various domains, providing robust solutions by exploring diverse behavioral niches. However, their application has primarily focused on static problems, with limited exploration in the context of dynamic combinatorial optimization problems. Furthermore, the theoretical understanding of QD algorithms remains underdeveloped, particularly when applied to learning heuristics instead of directly learning solutions in complex and dynamic combinatorial optimization domains, which introduces additional challenges. This paper introduces a novel QD framework for dynamic scheduling problems. We propose a map-building strategy that visualizes the solution space by linking heuristic genotypes to their behaviors, enabling their representation on a QD map. This map facilitates the discovery and maintenance of diverse scheduling heuristics. Additionally, we conduct experiments on both fixed and dynamically changing training instances to demonstrate how the map evolves and how the distribution of solutions unfolds over time. We also discuss potential future research directions that could enhance the learning process and broaden the applicability of QD algorithms to dynamic combinatorial optimization challenges.

Modern neural architecture search (NAS) is inherently multi-objective balancing trade-offs such as accuracy, parameter count, and computational cost. This complexity makes NAS computationally expensive and nearly impossible to solve without efficient approximations. To address this, we propose a novel surrogate modelling approach that leverages an ensemble of Siamese network blocks to predict dominance relationships between candidate architectures. Lightweight and easy to train, the surrogate achieves 92% accuracy and replaces the crowding distance calculation in the survivor selection strategy with a heuristic rule based on model size. Integrated into a framework termed SiamNAS, this design eliminates costly evaluations during the search process. Experiments on NAS-Bench-201 demonstrate the framework's ability to identify Pareto-optimal solutions with significantly reduced computational costs. The proposed SiamNAS identified a final non-dominated set containing the best architecture in NAS-Bench-201 for CIFAR-10 and the second-best for ImageNet, in terms of test error rate, within 0.01 GPU days. This proof-of-concept study highlights the potential of the proposed Siamese network surrogate model to generalise to multi-tasking optimisation, enabling simultaneous optimisation across tasks. Additionally, it offers opportunities to extend the approach for generating Sets of Pareto Sets (SOS), providing diverse Pareto-optimal solutions for heterogeneous task settings.

This paper introduces the Augmented Decision Space Optimization (ADSO) method for sparsity-driven optimization of mixed-strategies in Stackelberg Security Games (SSGs). The proposed method enhances traditional strategy optimization by combining binary variables to represent the presence of pure strategies with real-valued variables to refine their selection probabilities. Specifically, instead of waiting for an evolutionary process to gradually discover sparse solutions, the binary variables in ADS allow the real-valued variables to be switched on or off, thereby directly enforcing sparsity. This dual codification scheme achieves targets such as sparsification and computational efficiency in large-scale games. We demonstrate that ADS outperforms existing heuristic methods, offering superior solution quality, scalability, and stability. Empirical results across three different benchmark games show that ADS generates compact strategies with minimal computational overhead, achieving performance close to the exact methods. Furthermore, state-of-the-art results are obtained for problems where exact methods fail to scale effectively. Our framework promises broad applicability beyond SSGs, encompassing a wide range of game-theoretic and combinatorial optimization problems.

Despite the rapid progress in image generation, emotional image editing remains under-explored. The semantics, context, and structure of an image can evoke emotional responses, making emotional image editing techniques valuable for various real-world applications, including treatment of psychological disorders, commercialization of products, and artistic design. First, we present a novel challenge of emotion-evoked image generation, aiming to synthesize images that evoke target emotions while retaining the semantics and structures of the original scenes. To address this challenge, we propose a diffusion model capable of effectively understanding and editing source images to convey desired emotions and sentiments. Moreover, due to the lack of emotion editing datasets, we provide a unique dataset consisting of 340,000 pairs of images and their emotion annotations. Furthermore, we conduct human psychophysics experiments and introduce a new evaluation metric to systematically benchmark all the methods. Experimental results demonstrate that our method surpasses all competitive baselines. Our diffusion model is capable of identifying emotional cues from original images, editing images that elicit desired emotions, and meanwhile, preserving the semantic structure of the original images. All code, model, and dataset are available at GitHub.

Current zero-shot anomaly detection (ZSAD) methods show remarkable success in prompting large pre-trained vision-language models to detect anomalies in a target dataset without using any dataset-specific training or demonstration. However, these methods often focus on crafting/learning prompts that capture only coarse-grained semantics of abnormality, e.g., high-level semantics like "damaged", "imperfect", or "defective" objects. They therefore have limited capability in recognizing diverse abnormality details that deviate from these general abnormal patterns in various ways. To address this limitation, we propose FAPrompt, a novel framework designed to learn Fine-grained Abnormality Prompts for accurate ZSAD. To this end, a novel Compound Abnormality Prompt learning (CAP) module is introduced in FAPrompt to learn a set of complementary, decomposed abnormality prompts, where abnormality prompts are enforced to model diverse abnormal patterns derived from the same normality semantic. On the other hand, the fine-grained abnormality patterns can be different from one dataset to another. To enhance the cross-dataset generalization, another novel module, namely Data-dependent Abnormality Prior learning (DAP), is introduced in FAPrompt to learn a sample-wise abnormality prior from abnormal features of each test image to dynamically adapt the abnormality prompts to individual test images. Comprehensive experiments on 19 real-world datasets, covering both industrial defects and medical anomalies, demonstrate that FAPrompt substantially outperforms state-of-the-art methods in both image- and pixel-level ZSAD tasks. Code is available at https://github.com/mala-lab/FAPrompt.

Vital signs, such as heart rate and blood pressure, are critical indicators of patient health and are widely used in clinical monitoring and decision-making. While deep learning models have shown promise in forecasting these signals, their deployment in healthcare remains limited in part because clinicians must be able to trust and interpret model outputs. Without reliable uncertainty quantification - particularly calibrated prediction intervals (PIs) - it is unclear whether a forecasted abnormality constitutes a meaningful warning or merely reflects model noise, hindering clinical decision-making. To address this, we present two methods for deriving PIs from the Reconstruction Uncertainty Estimate (RUE), an uncertainty measure well-suited to vital-sign forecasting due to its sensitivity to data shifts and support for label-free calibration. Our parametric approach assumes that prediction errors and uncertainty estimates follow a Gaussian copula distribution, enabling closed-form PI computation. Our non-parametric approach, based on k-nearest neighbours (KNN), empirically estimates the conditional error distribution using similar validation instances. We evaluate these methods on two large public datasets with minute- and hour-level sampling, representing high- and low-frequency health signals. Experiments demonstrate that the Gaussian copula method consistently outperforms conformal prediction baselines on low-frequency data, while the KNN approach performs best on high-frequency data. These results underscore the clinical promise of RUE-derived PIs for delivering interpretable, uncertainty-aware vital sign forecasts.

Guided diffusion-model generation is a promising direction for customizing the generation process of a pre-trained diffusion model to address specific downstream tasks. Existing guided diffusion models either rely on training the guidance model with pre-collected datasets or require the objective functions to be differentiable. However, for most real-world tasks, offline datasets are often unavailable, and their objective functions are often not differentiable, such as image generation with human preferences, molecular generation for drug discovery, and material design. Thus, we need an **online** algorithm capable of collecting data during runtime and supporting a **black-box** objective function. Moreover, the **query efficiency** of the algorithm is also critical because the objective evaluation of the query is often expensive in real-world scenarios. In this work, we propose a novel and simple algorithm, **Fast Direct**, for query-efficient online black-box target generation. Our Fast Direct builds a pseudo-target on the data manifold to update the noise sequence of the diffusion model with a universal direction, which is promising to perform query-efficient guided generation. Extensive experiments on twelve high-resolution ($\small {1024 \times 1024}$) image target generation tasks and six 3D-molecule target generation tasks show $\textbf{6}\times$ up to $\textbf{10}\times$ query efficiency improvement and $\textbf{11}\times$ up to $\textbf{44}\times$ query efficiency improvement, respectively.

Vision-Language Models (VLMs) such as CLIP excel at zero-shot classification due to large-scale pre-training but are vulnerable to adversarial examples. Adversarial fine-tuning robustifies zero-shot models by aligning prediction scores of individual adversaries with their clean counterparts, which typically overlooks intermediate adversarial samples along the adversarial trajectory crossing the decision boundary. Such intermediate adversaries and their vicinity produce informative representations capturing the decision boundary in detail. They can be improved by sampling adversarial candidates from simplices formed by joining two consecutive vertices on the adversarial trajectory and their clean counterpart. However, sampling simplices for adversaries is very costly. To train robust VLM, we overcome these limitations by Taylor expansion and formulating an upper-bound of alignment loss that depends on the Jacobian/Hessian obtained at clean samples. As regions between clean and intermediate adversarial samples capture a larger decision landscape, we robustify VLM by plausible adversaries from simplices by our closed-form formulation equivalent to infinite uniform sampling of the simplex. We obtain state-of-the-art robustness across 15 datasets and diverse vision-language tasks.

Large language models (LLMs) have shown great potential as general-purpose AI assistants across various domains. To fully leverage this potential in specific applications, many companies provide fine-tuning API services, enabling users to upload their own data for LLM customization. However, fine-tuning services introduce a new safety threat: user-uploaded data, whether harmful or benign, can break the model’s alignment, leading to unsafe outputs. Moreover, existing defense methods struggle to address the diversity of fine-tuning datasets (e.g., varying sizes, tasks), often sacrificing utility for safety or vice versa. To address this issue, we propose Safe Delta, a safety-aware post-training defense method that adjusts the delta parameters (i.e., the parameter change before and after fine-tuning). Specifically, Safe Delta estimates the safety degradation, selects delta parameters to maximize utility while limiting overall safety loss, and applies a safety compensation vector to mitigate residual safety loss. Through extensive experiments on four diverse datasets with varying settings, our approach consistently preserves safety while ensuring that the utility gain from benign datasets remains unaffected.

Multi-objective optimization (MOO) exists extensively in machine learning, and aims to find a set of Pareto-optimal solutions, called the Pareto front, e.g., it is fundamental for multiple avenues of research in federated learning (FL). Pareto-Front Learning (PFL) is a powerful method implemented using Hypernetworks (PHNs) to approximate the Pareto front. This method enables the acquisition of a mapping function from a given preference vector to the solutions on the Pareto front. However, most existing PFL approaches still face two challenges: (a) sampling rays in high-dimensional spaces; (b) failing to cover the entire Pareto Front which has a convex shape. Here, we introduce a novel PFL framework, called as PHN-HVVS, which decomposes the design space into Voronoi grids and deploys a genetic algorithm (GA) for Voronoi grid partitioning within high-dimensional space. We put forward a new loss function, which effectively contributes to more extensive coverage of the resultant Pareto front and maximizes the HV Indicator. Experimental results on multiple MOO machine learning tasks demonstrate that PHN-HVVS outperforms the baselines significantly in generating Pareto front. Also, we illustrate that PHN-HVVS advances the methodologies of several recent problems in the FL field. The code is available at https://github.com/buptcmm/phnhvvs.

Solving various types of vehicle routing problems (VRPs) using a unified neural solver has garnered significant attentions in recent years. Despite their effectiveness, existing neural multi-task solvers often fail to account for the geometric structures inherent in different tasks, which may result in suboptimal performance. To address this limitation, we propose a curvature-aware pre-training framework. Specifically, we leverage mixed-curvature spaces during the feature fusion stage, encouraging the model to capture the underlying geometric properties of each instance. Through extensive experiments, we evaluate the proposed pre-training strategy on existing neural multi-task solvers across a variety of testing scenarios. The results demonstrate that the curvature-aware pre-training approach not only enhances the generalization capabilities of existing neural VRP solvers on synthetic datasets but also improves solution quality on real-world benchmarks.

Imitation learning (IL) has shown promise in various applications (e.g. robot locomotion) but is often limited to learning a single expert policy, constraining behavior diversity and robustness in unpredictable real-world scenarios. To address this, we introduce Quality Diversity Inverse Reinforcement Learning (QD-IRL), a novel framework that integrates quality-diversity optimization with IRL methods, enabling agents to learn diverse behaviors from limited demonstrations. This work introduces Extrinsic Behavioral Curiosity (EBC), which allows agents to receive additional curiosity rewards from an external critic based on how novel the behaviors are with respect to a large behavioral archive. To validate the effectiveness of EBC in exploring diverse locomotion behaviors, we evaluate our method on multiple robot locomotion tasks. EBC improves the performance of QD-IRL instances with GAIL, VAIL, and DiffAIL across all included environments by up to 185%, 42%, and 150%, even surpassing expert performance by 20% in Humanoid. Furthermore, we demonstrate that EBC is applicable to Gradient-Arborescence-based Quality Diversity Reinforcement Learning (QD-RL) algorithms, where it substantially improves performance and provides a generic technique for learning behavioral diverse policies. The source code of this work is provided at https://github.com/vanzll/EBC.

This paper presents an evolvable conditional diffusion method such that black-box, non-differentiable multi-physics models, as are common in domains like computational fluid dynamics and electromagnetics, can be effectively used for guiding the generative process to facilitate autonomous scientific discovery. We formulate the guidance as an optimization problem where one optimizes for a desired fitness function through updates to the descriptive statistic for the denoising distribution, and derive an evolution-guided approach from first principles through the lens of probabilistic evolution. Interestingly, the final derived update algorithm is analogous to the update as per common gradient-based guided diffusion models, but without ever having to compute any derivatives. We validate our proposed evolvable diffusion algorithm in two AI for Science scenarios: the automated design of fluidic topology and meta-surface. Results demonstrate that this method effectively generates designs that better satisfy specific optimization objectives without reliance on differentiable proxies, providing an effective means of guidance-based diffusion that can capitalize on the wealth of black-box, non-differentiable multi-physics numerical models common across Science.

Grounding open-domain knowledge from large language models (LLMs) into real-world reinforcement learning (RL) tasks represents a transformative frontier in developing intelligent agents capable of advanced reasoning, adaptive planning, and robust decision-making in dynamic environments. In this paper, we introduce the LLM-RL Grounding Taxonomy, a systematic framework that categorizes emerging methods for integrating LLMs into RL systems by bridging their open-domain knowledge and reasoning capabilities with the task-specific dynamics, constraints, and objectives inherent to real-world RL environments. This taxonomy encompasses both training-free approaches, which leverage the zero-shot and few-shot generalization capabilities of LLMs without fine-tuning, and fine-tuning paradigms that adapt LLMs to environment-specific tasks for improved performance. We critically analyze these methodologies, highlight practical examples of effective knowledge grounding, and examine the challenges of alignment, generalization, and real-world deployment. Our work not only illustrates the potential of LLM-RL agents for enhanced decision-making, but also offers actionable insights for advancing the design of next-generation RL systems that integrate open-domain knowledge with adaptive learning.

The rise of reinforcement learning (RL) in critical real-world applications demands a fundamental rethinking of privacy in AI systems. Traditional privacy frameworks, designed to protect isolated data points, fall short for sequential decision-making systems where sensitive information emerges from temporal patterns, behavioral strategies, and collaborative dynamics. Modern RL paradigms, such as federated RL (FedRL) and RL with human feedback (RLHF) in large language models (LLMs), exacerbate these challenges by introducing complex, interactive, and context-dependent learning environments that traditional methods do not address. In this position paper, we argue for a new privacy paradigm built on four core principles: multi-scale protection, behavioral pattern protection, collaborative privacy preservation, and context-aware adaptation. These principles expose inherent tensions between privacy, utility, and interpretability that must be navigated as RL systems become more pervasive in high-stakes domains like healthcare, autonomous vehicles, and decision support systems powered by LLMs. To tackle these challenges, we call for the development of new theoretical frameworks, practical mechanisms, and rigorous evaluation methodologies that collectively enable effective privacy protection in sequential decision-making systems.

Cognitive support in aviation operations can help reduce cognitive workload and improve flight safety. Existing data-driven models, typically trained for specific aircraft types, often lack generalization capabilities and cannot be directly applied to other aircrafts. The multimodal large language models (MLLMs) have shown significant value in various applications without additional training. Motivated by the success of MLLMs, we explore and develop a proactive aviation cognitive support framework (PACS) based on MLLMs to proactively assist pilots in anomaly perception and decision-making during aviation emergencies. We propose a localization-augmented anomaly perception method to address the limitations of MLLMs in recognizing small-sized warning message text. To improve the accuracy and efficiency in decision-making, we design a hierarchically structured aviation knowledge base and a "Retrieval-Selection Generation" strategy to generate accurate operational instructions. The experimental results demonstrate high accuracy in anomaly perception (87.51%) and decision-making (93.04%), while reducing token count by up to 56.77%. Our PACS is expected to adapt to various scenarios without additional fine-tuning, offering a new paradigm for future human-AI collaboration in aviation operations.

Text-to-3D shape generation has shown great promise in generating novel 3D content based on given text prompts. However, existing generative methods mainly consider geometric or visual plausibility while ignoring functionality for the generated 3D shapes. This greatly hinders the practicality of generated 3D shapes in real-world applications. Towards physical AI, we propose Fun3D, a physics-compliant functional text-to-3D shape generation method. By analyzing the solid mechanics of generated 3D shapes, we reveal that the 3D shapes generated by existing text-to-3D generation methods are impractical for real-world applications, as the generated 3D shapes do not comply with the physical laws. To this end, we leverage 3D diffusion models to provide 3D shape priors and design a data-driven differentiable physics layer to optimize 3D shape priors with solid mechanics. This allows us to optimize geometry efficiently and learn physical information about 3D shapes at the same time. Experimental results demonstrate that our method can consider both geometric plausibility and functional requirement, further bridging 3D virtual modeling and physical worlds to advance physical AI.

Contemporary Vision-Language Models (VLMs) such as CLIP offer an attractive zero-shot classification functionality facilitated by large-scale vision-language pre-training. However, they remain vulnerable to adversarial attacks, a critical security threat in realistic deployment. Adversarially robust fine-tuning provides generalizable robustness on new datasets while preserving natural performance by fine-tuning the pre-trained models. Fine-tuning robust CLIP typically relies on adversaries generated solely from the vision branch. However, this singular focus on the vision modality, coupled with static text prompts used as fixed category prototypes, limits the robustness achieved through dual-modality fine-tuning. We observe for CLIP fine-tuning that zero-shot adversarial robustness improves when we (i) stabilize the modality gap (a phenomenon where image and text features occupy different feature space regions) and (ii) lower/stabilize gradient norms. Both these steps enjoy further improvement of robustness if one fine-tunes with both visual and text adversaries. For both modalities, we leverage (i) the maximization of an effective rank of features and (ii) noise modulation of features. We show that maximizing the effective rank helps lower and stabilize the modality gap over adversaries with varying perturbation radii. The noise modulation of features, achieved by the so-called count sketching, lowers/stabilizes gradient norms. We outperform the state of the art on 15 datasets. We provide the first insights into the effects of modality gap & gradient norms in VLM fine-tuning.

Crowd simulation is crucial for urban planning, traffic management, public safety, and immersive environments. A fundamental challenge is capturing adaptive human behaviors that evolve dynamically with social interactions and task demands. Recently, physics-informed neural networks (PINNs) seamlessly integrate interpretable physics-based models with flexible data-driven learning, significantly enhancing simulation realism. However, current PINN-based methods typically rely on rigid representations of pedestrian perceptions and static task priorities of motion planning, limiting their ability to capture real-world social complexities and behavioral adaptability. To this end, we introduce SA-PINN, a novel Self-Adaptive Physics-Informed Neural Network specifically designed for modeling adaptive crowd behaviors. SA-PINN features two innovative adaptive modules: a self-adaptive social perception module, guided by a visual-field physics model to capture context-dependent social interactions dynamically; and a self-adaptive multi-task PINN training module, automatically balancing key motion objectives such as goal-reaching, collision avoidance, and alignment with real data. By jointly enabling perception-level and task-level adaptations within a unified physics-informed framework, SA-PINN generates highly realistic and physically consistent crowd simulations across diverse environmental contexts. Comprehensive evaluations on three real-world datasets (Lane, Cross 90, and GC) reveal that SA-PINN achieves a 29.7% gain in microscopic trajectory accuracy and enhances macroscopic density similarity by 23.5% compared to the best-performing baselines.

Long-horizon planning in robotic manipulation tasks requires translating underspecified, symbolic goals into executable control programs satisfying spatial, temporal, and physical constraints. However, language model-based planners often struggle with long-horizon task decomposition, robust constraint satisfaction, and adaptive failure recovery. We introduce InstructFlow, a multi-agent framework that establishes a symbolic, feedback-driven flow of information for code generation in robotic manipulation tasks. InstructFlow employs a InstructFlow Planner to construct and traverse a hierarchical instruction graph that decomposes goals into semantically meaningful subtasks, while a Code Generator generates executable code snippets conditioned on this graph. Crucially, when execution failures occur, a Constraint Generator analyzes feedback and induces symbolic constraints, which are propagated back into the instruction graph to guide targeted code refinement without regenerating from scratch. This dynamic, graph-guided flow enables structured, interpretable, and failure-resilient planning, significantly improving task success rates and robustness across diverse manipulation benchmarks, especially in constraint-sensitive and long-horizon scenarios.

Numerous well-established studies have demonstrated the superhuman capabilities of modern Vision-Language Models (VLMs) across a wide range of tasks. However, growing is the doubt about the continuing availability of reliable high-quality labeling (supervision) from human annotators, leading to stagnation of the model's performance. To address this challenge, ``superalignment'' employs the so-called weak-to-strong generalization paradigm, where the supervision from a weak model can provide generalizable knowledge for a strong model. While effective in aligning knowledge for clean samples between the strong and weak models, the standard weak-to-strong approach typically fails to capture adversarial robustness, exposing strong VLMs to adversarial attacks. This inability to transfer adversarial robustness is because adversarial samples are normally missing in the superalignment stage. To this end, we are the first to propose the weak-to-strong (adversarial) robustness generalization method to elicit zero-shot robustness in large-scale models by an unsupervised scheme, mitigating the unreliable information source for alignment from two perspectives: alignment re-weighting and source guidance refinement. We analyze settings under which robustness generalization is possible. Extensive experiments across various vision-language benchmarks validate the effectiveness of our method in numerous scenarios, demonstrating its plug-and-play applicability to large-scale VLMs.

As foundation Vision-Language Models (VLMs) unlock fine-tuning on smaller datasets while leveraging large-scale pre-training data, machine unlearning becomes critical in addressing privacy concerns and regulatory compliance. Task vector, representing the difference between parameters of models fine-tuned with and without specific data, is a popular retraining-free unlearning strategy. However, we observe that task vectors exhibit substantial sensitivity to various fine-tuning configurations, resulting in unstable unlearning effectiveness that correlates negatively with the prediction-level variance. While aggregating multiple functions (e.g., VLM with classifier) whose parameters are represented by different task vectors reduces function variance and improves unlearning, the computational cost of obtaining numerous task vectors and aggregating functions is computationally high. Thus, in order to capture the space of task vectors induced by diverse fine-tuning strategies, we propose modeling it within the convex hull of $(Q-1)$-simplex whose vertices represent $Q$ task vectors. Although a function ensemble can be formed by sampling numerous task vectors from such a simplex, we derive a closed-form ensemble of an infinite number of functions whose parameters are uniformly sampled from the simplex, enabling efficient function-level task vector ensembling with enhanced unlearning performance. Extensive experiments and analyses across diverse datasets and scenarios demonstrate the efficacy of our method.

Semi-Unbalanced Optimal Transport (SemiUOT) shows great promise in matching two probability measures by relaxing one of the marginal constraints. Previous solvers often incorporate an entropy regularization term, which can result in inaccurate matching solutions. To address this issue, we focus on determining the marginal probability distribution of SemiUOT with KL divergence using the proposed Equivalent Transformation Mechanism (ETM) approach. Furthermore, we extend the ETM-based method into exploiting the marginal probability distribution of Unbalanced Optimal Transport (UOT) with KL divergence for validating its generalization. Once the marginal probabilities of UOT/SemiUOT are determined, they can be transformed into a classical Optimal Transport (OT) problem. Moreover, we propose a KKT-Multiplier regularization term combined with Multiplier Regularized Optimal Transport (MROT) to achieve more accurate matching results. We conduct several numerical experiments to demonstrate the effectiveness of our proposed methods in addressing UOT/SemiUOT problems.

Social recommendations (SRs) aim to enhance preference modeling by integrating social networks. However, their effectiveness is mainly constrained by two factors: the noisy social connections that may not reflect shared interests, and the limited number of social connections for most users, which hampers the system's ability to fully leverage social influence. Therefore, it is essential to perform social network refinement by removing noisy connections and adding meaningful ones for robust SRs. Inspired by the denoising capability of generative diffusion models, we propose a Model-Agnostic Social Network Refinement framework with Diffusion Models for Robust Social Recommendation (ARD-SR). Specifically, in the forward process, we corrupt the social network by progressively adding position-specific Gaussian noise calibrated to the user preference similarity, better simulating how the social network responds to noise perturbations. The reverse process learns to denoise, guided by each user's neighborhood preferences from the SR backbone, generating a tailored social network aligned with each user's preference for establishing connections. For effective learning, we design a curriculum-based training mechanism that progressively introduces challenging samples characterized by high sparsity or high noise levels. Finally, ARD-SR and the SR backbone are alternately trained, ensuring a continuous mutual enhancement between the social network refinement and the backbone's user representation learning. To further enhance the quality of the refined social network, (1) we introduce a preference-guided flip operation during inference to improve the input quality; and (2) we modify social connections based on the exponential weighted moving average of ARD-SR's predictions across epochs to reduce fluctuations. Experiments on three datasets show that ARD-SR significantly improves SR performance across multiple SR backbones. The code is released at https://github.com/sunyc123r/ARD-SR.

To reconstruct the 3D geometry from calibrated images, learning-based multi-view stereo (MVS) methods typically perform multi-view depth estimation and then fuse depth maps into a mesh or point cloud. To improve the computational efficiency, many methods initialize a coarse depth map and then gradually refine it in higher resolutions. Recently, diffusion models achieve great success in generation tasks. Starting from a random noise, diffusion models gradually recover the sample with an iterative denoising process. In this paper, we propose a novel MVS framework, which introduces diffusion models in MVS. Specifically, we formulate depth refinement as a conditional diffusion process. Considering the discriminative characteristic of depth estimation, we design a condition encoder to guide the diffusion process. To improve efficiency, we propose a novel diffusion network combining lightweight 2D U-Net and convolutional GRU. Moreover, we propose a novel confidence-based sampling strategy to adaptively sample depth hypotheses based on the confidence estimated by diffusion model. Based on our novel MVS framework, we propose two novel MVS methods, DiffMVS and CasDiffMVS. DiffMVS achieves competitive performance with state-of-the-art efficiency in run-time and GPU memory. CasDiffMVS achieves state-of-the-art performance on DTU, Tanks & Temples and ETH3D.

Real-time solutions to the influence blocking maximization (IBM) problems are crucial for promptly containing the spread of misinformation. However, achieving this goal is nontrivial, mainly because assessing the blocked influence of an IBM problem solution typically requires plenty of expensive Monte Carlo simulations (MCSs). This work presents a novel approach that enables solving IBM problems with hundreds of thousands of nodes and edges in seconds. The key idea is to construct a fast-to-evaluate surrogate model called neural influence estimator (NIE) offline as a substitute for the time-intensive MCSs, and then combine it with optimization algorithms to address IBM problems online. To this end, a learning problem is formulated to build the NIE that takes the false-and-true information instance as input, extracts features describing the topology and interrelationship between two seed sets, and predicts the blocked influence. A well-trained NIE can generalize across different IBM problems given a social network, and can be readily combined with existing IBM optimization algorithms. The experiments on 25 IBM problems with up to millions of edges show that the NIE-based optimization method can be up to four orders of magnitude faster than MCSs-based optimization method to achieve the same optimization quality. Moreover, given a one-minute limit, the NIE-based method can solve IBM problems with up to hundreds of thousands of nodes, which is at least one order of magnitude larger than what can be solved by existing methods.

In the field of multiagent reinforcement learning (MARL), the ability to effectively explore unknown environments and collect information and experiences that are most beneficial for policy learning represents a critical research area. However, existing work often encounters difficulties in addressing the uncertainties caused by state changes and the inconsistencies between agents' local observations and global information, which presents significant challenges to coordinated exploration among multiple agents. To address this issue, this article proposes a novel MARL exploration method with Min-Max intrinsic motivation (E2M) that promotes the learning of joint policies of agents by introducing surprise minimization and social influence maximization. Since the agent is subject to unstable state changes in the environment, we introduce surprise minimization by computing state entropy to encourage the agents to cope with more stable and familiar situations. This method enables surprise estimation based on the low-dimensional representation of states obtained from random encoders. Furthermore, to prevent surprise minimization from leading to conservative policies, we introduce mutual information between agents' behaviors as social influence. By maximizing social influence, the agents are encouraged to interact to facilitate the emergence of cooperative behavior. The performance of our proposed E2M is testified across a range of popular StarCraft II and Multiagent MuJoCo tasks. Comprehensive results demonstrate its effectiveness in enhancing the cooperative capability of the multiple agents.

Transfer multiobjective optimization promises sample-efficient discovery of near Pareto-optimal solutions to a target task by utilizing experiential priors from related source tasks. In this paper, we show that in domains where evaluation data is at a premium, e.g., in scientific and engineering disciplines involving time-consuming computer simulations or complex real-world experimentation, knowledge transfer through surrogate models can be pivotal in saving sample evaluation costs. While state-of-the-art algorithms (without transfer) typically assume budgets in the order of only a few hundred evaluations, we seek to explore how far we can get on even tighter budgets. The uniqueness of our proposed Expensive Transfer Evolutionary Multiobjective Optimizer (ExTrEMO) is that it can maximally utilize external information from hundreds of source datasets, including those that may be negatively correlated with the target task. This is achieved by melding evolutionary search with factorized transfer Gaussian process surrogates, capturing varied source-target correlations in potentially decentralized computation environments. We provide a regret bound analysis for ExTrEMO that translates to a theoretical proof of increasingly faster convergence as a result of multi-source transfers. The theory is experimentally verified on benchmark functions and toward accelerated design of biomedical microdevices. We release our code at https://github.com/LiuJ-2023/ExTrEMO.

High-dimensional and incomplete (HDI) data are frequently encountered in diverse real-world applications involving complex interactions among numerous nodes. Approaches based on latent feature analysis (LFA) have proven effective in performing representation learning in HDI data. Nevertheless, they cannot handle the high-order connectivity among nodes in HDI data well, resulting in severe accuracy loss. To address the previously mentioned issue, we present a novel model in this paper, namely Graph Linear Convolution Pooling Network (GLCPN). The proposed GLCPN adopts the three-fold ideas. First, it leverages simplified graph convolutions to efficiently capture high-order connectivity among nodes for learning representations of matrix factorization. Second, a simple yet effective priori convolution operator is adopted by each graph neural layer to capture node-node collaboration for aggregation. Third, a locality-enhanced pooling scheme is designed to holistically utilize multi-layer representations of the neighborhood. Therefore, GLCPN can effectively acquire the hidden information in HDI data with high efficiency. In addition, we have conducted a theoretical analysis demonstrating that the proposed GLCPN is more expressive compared with existing graph neural networks for HDI data. Extensive experiments have been further conducted on ten well-established HDI datasets from various applications. The experimental results demonstrate that the proposed GLCPN significantly outperforms state-of-the-art models for learning representations in HDI data evaluated by accuracy and efficiency metrics.

The rapid growth of the internet has led to an alarming increase in the dissemination of fake news, which has had many negative effects on society. Various methods have been proposed for detecting fake news. However, these approaches suffer from several limitations. First, most existing works only consider news as separate entities and do not consider the correlations between fake news and real news. Moreover, these works are usually conducted in the Euclidean space, which is unable to capture complex relationships between news, in particular the hierarchical relationships. To tackle these issues, we introduce a novelMulti-modalHyperbolicRepresentation framework (MHR) for fake news detection. Specifically, we capture the correlations between news for graph construction to arrange and analyze different news. To fully utilize the multi-modal characteristics, we first extract the textual and visual information, and then design a Lorentzian multi-modal fusion module to fuse them as the node information in the graph. By utilizing the fully hyperbolic graph neural networks, we learn the graph’s representation in hyperbolic space, followed by a detector for detecting fake news. The experimental results on three real-world datasets demonstrate that our proposed MHR model achieves state-of-the-art performance, indicating the benefits of hyperbolic representation.

Diverse human motion prediction (HMP) aims to predict multiple plausible future motions given an observed human motion sequence. It is a challenging task due to the diversity of potential human motions while ensuring an accurate description of future human motions. Current solutions are either low-diversity or limited in expressiveness. Recent denoising diffusion probabilistic models (DDPM) demonstrate promising performance in various generative tasks. However, introducing DDPM directly into diverse HMP incurs some issues. While DDPM can enhance the diversity of potential human motion patterns, the predicted human motions gradually become implausible over time due to significant noise disturbances in the forward process of DDPM. This phenomenon leads to the predicted human motions being unrealistic, seriously impacting the quality of predicted motions and restricting their practical applicability in real-world scenarios. To alleviate this, we propose a novel conditional diffusion-based generative model, called DivDiff, to predict more diverse and realistic human motions. Specifically, the DivDiff employs DDPM as our backbone and incorporates Discrete Cosine Transform (DCT) and Transformer mechanisms to encode the observed human motion sequence as a condition to instruct the reverse process of DDPM. More importantly, we design a diversified reinforcement sampling function (DRSF) to enforce human skeletal constraints on the predicted human motions. DRSF utilizes the acquired information from human skeletal as prior knowledge, thereby reducing significant disturbances introduced during the forward process. Extensive results received in the experiments on two widely-used datasets (Human3.6M and HumanEva-I) demonstrate that our model obtains competitive performance on both diversity and accuracy.

In this article, we present the dynamic graph mixer (DGM), a novel model for learning spatiotemporal-individual coupled features from high-dimensional and incomplete (HDI) tensors, which frequently represent dynamic interactions among real-world data samples. In contrast to existing methods, the proposed DGM possesses the following three advantages when learning representations from HDI tensors. First, it performs light graph message passing based on the conjoint attentions learned by jointly modeling latent features and implicit structures to extract the high-order connectivity. Second, a multilayer nonlinear tensor neural network (TNN) is adopted to learn the intricate attribute features of node-node-time from different views. Third, it follows the Tucker decomposition paradigm in a data density-oriented modeling mechanism to integrate node representations, preserving the overall multidimensional interaction patterns. In addition, we provide theoretical evidence that the key components in DGM can significantly improve expressiveness. Extensive experiments conducted on eight testing datasets of HDI tensors demonstrate that DGM outperforms state-of-the-art methods in both learning accuracy and efficiency.

Few-shot image classification (FSIC) is beneficial for a variety of real-world scenarios, aiming to construct a recognition system with limited training data. In this article, we extend the original FSIC task by incorporating defense against malicious adversarial examples. This can be an arduous challenge because numerous deep learning-based approaches remain susceptible to adversarial examples, even when trained with ample amounts of data. Previous studies on this problem have predominantly concentrated on the meta-learning framework, which involves sampling numerous few-shot tasks during the training stage. In contrast, we propose a straightforward but effective baseline via learning robust and discriminative representations without tedious meta-task sampling, which can further be generalized to unforeseen adversarial FSIC tasks. Specifically, we introduce an adversarial-aware (AA) mechanism that exploits feature-level distinctions between the legitimate and the adversarial domains to provide supplementary supervision. Moreover, we design a novel adversarial reweighting training strategy to ameliorate the imbalance among adversarial examples. To further enhance the adversarial robustness without compromising discriminative features, we propose the cyclic feature purifier during the postprocessing projection, which can reduce the interference of unforeseen adversarial examples. Furthermore, our method can obtain robust feature embeddings that maintain superior transferability, even when facing cross-domain adversarial examples. Extensive experiments and systematic analyses demonstrate that our method achieves state-of-the-art robustness as well as natural performance among adversarially robust FSIC algorithms on three standard benchmarks by a substantial margin.

With the rapid development of Internet and Web techniques, Cross-Domain Recommendation (CDR) models have been widely explored for resolving the data-sparsity and cold-start problem. Meanwhile, most CDR models should utilize explicit domain-shareable information (e.g., overlapped users or items) for knowledge transfer across domains. However, this assumption may not be always satisfied since users and items are always non-overlapped in real practice. The performance of many previous works will be severely impaired when these domain-shareable information are not available. To address the aforementioned issues, we propose the Joint Preference Exploration and Dynamic Embedding Transportation model (JPEDET) in this paper which is a novel framework for solving the CDR problem when users and items are non-overlapped. JPEDET includes two main modules, i.e., joint preference exploration module and dynamic embedding transportation module. The joint preference exploration module aims to fuse rating and review information for modelling user preferences. The dynamic embedding transportation module is set to share knowledge via neural ordinary equations for dual transformation across domains. Moreover, we innovatively propose the dynamic transport flow equipped with linear interpolation guidance on barycentric Wasserstein path for achieving accurate and bidirectional transformation. Our empirical study on Amazon datasets demonstrates that JPEDET significantly outperforms the state-of-the-art models under the CDR setting.

Video-based facial analysis is important for autonomous agents to understand human expressions and sentiments. However, limited labeled data is available to learn effective facial representations. This paper proposes a novel self-supervised face-centric pretraining framework, called PrefAce, which learns transferable video facial representation without labels. The self-supervised learning is performed with an effective landmark-guided global-local tube distillation. Meanwhile, a novel instance-wise update FaceFeat Cache is built to enforce more discriminative and diverse representations for downstream tasks. Extensive experiments demonstrate that the proposed framework learns universal instance-aware facial representations with fine-grained landmark details from videos. The point is that it can transfer across various facial analysis tasks, e.g., Facial Attribute Recognition (FAR), Facial Expression Recognition (FER), DeepFake Detection (DFD), and Lip Synchronization (LS). Our framework also outperforms the state-of-the-art on various downstream tasks, even in low data regimes. Code is available at https://github.com/siyuan-h/PrefAce.

Federated learning (FL) provides a privacy-preserving approach for collaborative training of machine learning models. Given the potential data heterogeneity, it is crucial to select appropriate collaborators for each FL participant (FL-PT) based on data complementarity. Recent studies have addressed this challenge. Similarly, it is imperative to consider the inter-individual relationships among FL-PTs where some FL-PTs engage in competition. Although FL literature has acknowledged the significance of this scenario, practical methods for establishing FL ecosystems remain largely unexplored. In this paper, we extend a principle from the balance theory, namely “the friend of my enemy is my enemy”, to ensure the absence of conflicting interests within an FL ecosystem. The extended principle and the resulting problem are formulated via graph theory and integer linear programming. A polynomial-time algorithm is proposed to determine the collaborators of each FL-PT. The solution guarantees high scalability, allowing even competing FL-PTs to smoothly join the ecosystem without conflict of interest. The proposed framework jointly considers competition and data heterogeneity. Extensive experiments on real-world and synthetic data demonstrate its efficacy compared to five alternative approaches, and its ability to establish efficient collaboration networks among FL-PTs.

Evolutionary algorithms (EAs) have achieved remarkable success in tackling complex combinatorial optimization problems. However, EAs often demand carefully-designed operators with the aid of domain expertise to achieve satisfactory performance. In this work, we present the first study on large language models (LLMs) as evolutionary combinatorial optimizers. The main advantage is that it requires minimal domain knowledge and human efforts, as well as no additional training of the model. This approach is referred to as LLM-driven EA (LMEA). Specifically, in each generation of the evolutionary search, LMEA instructs the LLM to select parent solutions from current population, and perform crossover and mutation to generate offspring solutions. Then, LMEA evaluates these new solutions and include them into the population for the next generation. LMEA is equipped with a self-adaptation mechanism that controls the temperature of the LLM. This enables it to balance between exploration and exploitation and prevents the search from getting stuck in local optima. We investigate the power of LMEA on the classical traveling salesman problems (TSPs) widely used in combinatorial optimization research. Notably, the results show that LMEA performs competitively to traditional heuristics in finding high-quality solutions on TSP instances with up to 20 nodes. Additionally, we also study the effectiveness of LLM-driven crossover/mutation and the self- adaptation mechanism in evolutionary search. In summary, our results reveal the great potentials of LLMs as evolutionary optimizers for solving combinatorial problems. We hope our research shall inspire future explorations on LLM-driven EAs for complex optimization challenges.

Recently, evolutionary multitasking has been employed to generate a “set of Pareto sets” (SOS) for machine learning models, addressing diverse task settings across heterogeneous environments. This involves creating a repository of compact, specialized solution models that are collectively tailored to each specific task setting and environment, enabling users to select the most suitable model based on particular specifications and preferences. In this paper, we further demonstrate the versatility and applicability of the SOS concept across diverse domains, focusing on three real-world problems: engineering design problems, inventory management problems, and hyperparameter optimization problems. Additionally, as evolutionary multitasking has proven effective in generating the SOS, we investigate the performance of current evolutionary multitasking methods on these real-world problems. Subsequently, we present visualizations of the generated SOS in both decision and objective spaces, complemented by the development of a measurement to gauge the similarity between different Pareto sets corresponding to diverse tasks. Finally, we show that by systematically examining the shifts in Pareto optimal designs across different task settings though the SOS solutions, users can gain deeper understandings on the dynamic interplay between design solutions and their performance in different settings or contexts.

Rapid advancements in text-to-3D shape synthesis using generative AI models have achieved remarkable general-ization performance, generating novel designs based on com-binations of different concepts. However, designs synthesized through automated optimization often result in ill-defined shapes, which render such designs impractical for engineering design applications. We present the first end-to-end prompt evolution design optimization (PREDO) framework contextualized in a vehicle design scenario that leverages a vision-language model to penalize impractical car designs synthesized by a generative model. The backbone of our framework is an evolutionary strategy coupled with an optimization objective function that comprises a physics-based solver and a vision-language model for practical or functional guidance in the generated car designs. In the prompt evolution search, the optimizer iteratively generates a population of text prompts, which embed user specifications on the aerodynamic performance and visual preferences of the 3D car designs. Then, in addition to the computational fluid dynamics simulations, the pre-trained vision-language model is used to penalize impractical designs and, thus, foster the evolutionary algorithm to seek more viable designs. Our investigations on a car design optimization problem show a wide spread of potential car designs generated at the early phase of the search, which indicates a good diversity of designs in the initial populations, and an increase of over 20 % in the probability of generating practical designs compared to a baseline framework without using a vision-language model. Visual inspection of the designs against the performance results demonstrates prompt evolution as a very promising paradigm for finding novel designs with good optimization performance while providing ease of use in specifying design specifications and preferences via a natural language interface.

Adversarially robust knowledge distillation aims to com-press large-scale models into lightweight models while preserving adversarial robustness and natural performance on a given dataset. Existing methods typically align probability distributions of natural and adversarial samples between teacher and student models, but they overlook intermediate adversarial samples along the “adversarial path” formed by the multi-step gradient ascent of a sample towards the decision boundary. Such paths capture rich information about the decision boundary. In this paper, we propose a novel adversarially robust knowledge distillation approach by incorporating such adversarial paths into the alignment process. Recognizing the diverse impacts of intermediate adversarial samples (ranging from benign to noisy), we propose an adaptive weighting strategy to selectively em-phasize informative adversarial samples, thus ensuring efficient utilization of lightweight model capacity. Moreover, we propose a dual-branch mechanism exploiting two following insights: (i) complementary dynamics of adversar-ial paths obtained by targeted and untargeted adversarial learning, and (ii) inherent differences between the gradient ascent path from class$c_{i}$towards the nearest class bound-ary and the gradient descent path from a specific class$c_{j}$towards the decision region of$c_{i}(i\neq j)$. Comprehensive experiments demonstrate the effectiveness of our method on lightweight models under various settings.

Few-shot learning (FSL) facilitates a variety of computer vision tasks yet remains vulnerable to adversarial attacks. Existing adversarially robust FSL methods rely on either visual similarity learning or class concept learning. Our analysis reveals that these two learning paradigms are complementary, exhibiting distinct robustness due to their unique decision boundary types (concepts clustering by the visual similarity label vs. classification by the class labels). To bridge this gap, we propose a novel framework unifying adversarially robust similarity learning and class concept learning. Specifically, we distill parameters from both network branches into a “unified embedding model” during robust optimization and redistribute them to individual network branches periodically. To capture generalizable robustness across diverse branches, we initialize adversaries in each episode with cross-branch class-wise “global adversarial perturbations” instead of less informative random initialization. We also propose a branch robustness harmonization to modulate the optimization of similarity and class concept learners via their relative adversarial robustness. Extensive experiments demonstrate the state-of-the-art performance of our method in diverse few-shot scenarios.

Weakly-supervised Spatio-Temporal Video Grounding(STVG) aims to localize target object tube given a text query, without densely annotated training data.Existing methods extract each candidate tube feature independently by cropping objects from video frame feature, discarding all contextual information such as position change and inter-entity relationship.In this paper, we propose Video-Text Prompting(VTP) to construct candidate feature.Instead of cropping tube region from feature map, we draw visual markers(e.g.red circle) over objects tubes as video prompts; corresponding text prompt(e.g. in red circle) is also inserted after the subject word of query text to highlight its presence.Nevertheless, each candidate feature may look similar without cropping.To address this, we further propose Contrastive VTP(CVTP) by introducing negative contrastive samples whose candidate object is erased instead of being highlighted; by comparing the difference between VTP candidate and the contrastive sample, the gap of matching score between correct candidate and the rest is enlarged.Extensive experiments and ablations are conducted on several STVG datasets and our results surpass existing weakly-supervised methods by a great margin, demonstrating the effectiveness of our proposed methods.

Cross-Domain Recommendation (CDR) have become increasingly appealing by leveraging useful information to tackle the data sparsity problem across domains. Most of latest CDR models assume that domain-shareable user-item information (e.g., rating and review on overlapped users or items) are accessible across domains. However, these assumptions become impractical due to the strict data privacy protection policy. In this paper, we propose Reducing Item Discrepancy (RidCDR) model on solving Privacy-Preserving Cross-Domain Recommendation (PPCDR) problem. Specifically, we aim to enhance the model performance on both source and target domains without overlapped users and items while protecting the data privacy. We innovatively propose private-robust embedding alignment module in RidCDR for knowledge sharing across domains while avoiding negative transfer privately. Our empirical study on Amazon and Douban datasets demonstrates that RidCDR significantly outperforms the state-of-the-art models under the PPCDR without overlapped users and items.

The next Point-of-interest recommendation has attracted extensive research interest recently, which predicts users' subsequent movements. The main challenge is how to effectively capture users' personalized sequential transitions in check-in trajectory, and various methods have been developed. However, most existing studies ignore the temporal information when conducting the next POI recommendation. To fill this gap, we investigate a time-specific next POI recommendation task, which additionally incorporates the target time information. We propose a brand new Time2Rotation technique to capture the temporal information. Different from conventional methods, we represent timeslots as rotation vectors and then perform the rotation operations. Based on the Time2Rotation technique, we propose a novel rotation-based temporal attention network, namely ROTAN, for the time-specific next POI recommendation task. The ROTAN begins by building a collaborative POI transition graph, capturing the asymmetric temporal influence in sequential transitions. After that, it incorporates temporal information into the modeling of individual check-in trajectories, extracting separate representations for user preference and POI influence to reflect their distinct temporal patterns. Lastly, the target time is integrated to generate recommendations. Extensive experiments are conducted on three real-world datasets, which demonstrates the advantages of the proposed Time2Rotation technique and ROTAN recommendation model.

Social Recommendation (SR) typically exploits neighborhood influence in the social network to enhance user preference modeling. However, users' intricate social behaviors may introduce noisy social connections for user modeling and harm the models' robustness. Existing solutions to alleviate social noise either filter out the noisy connections or generate new potential social connections. Due to the absence of labels, the former approaches may retain uncertain connections for user preference modeling while the latter methods may introduce additional social noise. Through data analysis, we discover that (1) social noise likely comes from the connected users with low preference similarity; and (2) Opinion Leaders (OLs) play a pivotal role in influence dissemination, surpassing high-similarity neighbors, regardless of their preference similarity with trusting peers. Guided by these observations, we propose a novel Self-Supervised Denoising approach through Independent Cascade Graph Augmentation, for more robust SR. Specifically, we employ the independent cascade diffusion model to generate an augmented graph view, which traverses the social graph and activates the edges in sequence to simulate the cascading influence spread. To steer the augmentation towards a denoised social graph, we (1) introduce a hierarchical contrastive loss to prioritize the activation of OLs first, followed by high-similarity neighbors, while weakening the low-similarity neighbors; and (2) integrate an information bottleneck based contrastive loss, aiming to minimize mutual information between original and augmented graphs yet preserve sufficient information for improved SR. Experiments conducted on two public datasets demonstrate that our model outperforms the state-of-the-art while also exhibiting higher robustness to different extents of social noise.

Long-term time series forecasting (LTSF) aims to predict future values for a long time based on historical data. The period term is an essential component of the time series, which is complex yet important for LTSF. Although existing studies have achieved promising results, they still have limitations in modeling dynamic complicated periods. Most studies only focus on static periods with fixed time steps, while very few studies attempt to capture dynamic periods in the time domain. In this paper, we dissect the original time series in time and frequency domains and empirically find that changes in periods are more easily captured and quantified in the frequency domain. Based on this observation, we propose to explore dynamic period features using rotation in the frequency domain. To this end, we develop the frequency-based rotation network (FRNet), a novel LTSF method to effectively capture the features of the dynamic complicated periods. FRNet decomposes the original time series into period and trend components. Based on the complex-valued linear networks, it leverages a period frequency rotation module to predict the period component and a patch frequency rotation module to predict the trend component, respectively. Extensive experiments on seven real-world datasets consistently demonstrate the superiority of FRNet over various state-of-the-art methods. The source code is available at https://github.com/SiriZhang45/FRNet.

Recent generative methods have revolutionized the way of human motion synthesis, such as Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), and Denoising Diffusion Probabilistic Models (DMs). These methods have gained significant attention in human motion fields. However, there are still challenges in unconditionally generating highly diverse human motions from a given distribution. To enhance the diversity of synthesized human motions, previous methods usually employ deep neural networks (DNNs) to train a transport map that transforms Gaussian noise distribution into real human motion distribution. According to Figalli's regularity theory, the optimal transport map computed by DNNs frequently exhibits discontinuities. This is due to the inherent limitation of DNNs in representing only continuous maps. Consequently, the generated human motions tend to heavily concentrate on densely populated regions of the data distribution, resulting in mode collapse or mode mixture. To address the issues, we propose an efficient method called MOOT for unconditional human motion synthesis. First, we utilize a reconstruction network based on GRU and transformer to map human motions to latent space. Next, we employ convex optimization to match the noise distribution with the latent space distribution of human motions through the Optimal Transport (OT) map. Then, we combine the extended OT map with the generator of reconstruction network to generate new human motions. Thereby overcoming the issues of mode collapse and mode mixture. MOOT generates a latent code distribution that is well-behaved and highly structured, providing a strong motion prior for various applications in the field of human motion. Through qualitative and quantitative experiments, MOOT achieves state-of-the-art results surpassing the latest methods, validating its superiority in unconditional human motion generation.

Federated learning (FL) is a machine learning paradigm that allows multiple FL participants (FL-PTs) to collaborate on training models without sharing private data. Due to data heterogeneity, negative transfer may occur in the FL training process. This necessitates FL-PT selection based on their data complementarity. In cross-silo FL, organizations that engage in business activities are key sources of FL-PTs. The resulting FL ecosystem has two features: (i) self-interest, and (ii) competition among FL-PTs. This requires the desirable FL-PT selection strategy to simultaneously mitigate the problems of free riders and conflicts of interest among competitors. To this end, we propose an optimal FL collaboration formation strategy -FedEgoists- which ensures that: (1) a FL-PT can benefit from FL if and only if it benefits the FL ecosystem, and (2) a FL-PT will not contribute to its competitors or their supporters. It provides an efficient clustering solution to group FL-PTs into coalitions, ensuring that within each coalition, FL-PTs share the same interest. We theoretically prove that the FL-PT coalitions formed are optimal since no coalitions can collaborate together to improve the utility of any of their members. Extensive experiments on widely adopted benchmark datasets demonstrate the effectiveness of FedEgoists compared to nine state-of-the-art baseline methods, and its ability to establish efficient collaborative networks in cross-silos FL with FL-PTs that engage in business activities.

Due to privacy and security concerns, recent advancements in group fairness advocate for model training regardless of demographic information. However, most methods still require prior knowledge of demographics. In this study, we explore the potential for achieving fairness without compromising its utility when no prior demographics are provided to the training set, namely _harmless Rawlsian fairness_. We ascertain that such a fairness requirement with no prior demographic information essential promotes training losses to exhibit a Dirac delta distribution. To this end, we propose a simple but effective method named VFair to minimize the variance of training losses inside the optimal set of empirical losses. This problem is then optimized by a tailored dynamic update approach that operates in both loss and gradient dimensions, directing the model towards relatively fairer solutions while preserving its intact utility. Our experimental findings indicate that regression tasks, which are relatively unexplored from literature, can achieve significant fairness improvement through VFair regardless of any prior, whereas classification tasks usually do not because of their quantized utility measurements. The implementation of our method is publicly available at https://github.com/wxqpxw/VFair.

Road network representation learning aims to learn compressed and effective vectorized representations for road segments that are applicable to numerous tasks. In this paper, we identify the limitations of existing methods, particularly their overemphasis on the distance effect as outlined in the First Law of Geography. In response, we propose to endow road network representation with the principles of the recent Third Law of Geography. To this end, we propose a novel graph contrastive learning framework that employs geographic configuration-aware graph augmentation and spectral negative sampling, ensuring that road segments with similar geographic configurations yield similar representations, and vice versa, aligning with the principles stated in the Third Law. The framework further fuses the Third Law with the First Law through a dual contrastive learning objective to effectively balance the implications of both laws. We evaluate our framework on two real-world datasets across three downstream tasks. The results show that the integration of the Third Law significantly improves the performance of road segment representations in downstream tasks.

Most existing denoising recommendation methods alleviate noisy implicit feedback (user behaviors) through mainly empirical studies. However, such studies may lack theoretical explainability and fail to model comprehensive noise patterns, which hinders the understanding and capturing of different noise patterns that affect users’ behaviors. Thus, we propose to capture comprehensive noise patterns through theoretical and empirical analysis for more effective denoising, where users’ behaviors are divided into willingness and action phases to disentangle independent noise patterns. Willingness refers to the user’s intent to interact with an item, which may not lead to actual interaction due to different factors such as misclicking. Action denotes the user’s actual interaction with an item. Our analysis unveils that (1) in the willingness phase, high uncertainty in the user’s willingness to interact with the item can lead to high expectation loss which aligns with the findings of existing denoising methods; and (2) in the action phase, higher user-specific inconsistency between willingness and action not only leads to more noise in the user’s overall behaviors but also makes it harder to distinguish between true and noisy behaviors. Inspired by these findings, we propose a Unified Denoising Training (UDT) method for recommendation. To alleviate uncertainty in the willingness phase, we lower the importance of the user-item interaction with high willingness uncertainty recognized by high loss. To ease the inconsistency in the action phase, we lower the importance for users with high user-specific inconsistency as it may lead to noisier behaviors. Then, we increase the importance gap between the clean and noisy behaviors for users with low user-specific inconsistency as their behaviors are more distinguishable. Extensive experiments on three real-world datasets show that our proposed UDT outperforms state-of-the-art denoising recommendation methods.

Most existing bundle generation approaches fall short in generating fixed-size bundles. Furthermore, they often neglect the underlying user intents reflected by the bundles in the generation process, resulting in less intelligible bundles. This paper addresses these limitations through the exploration of two interrelated tasks, i.e., personalized bundle generation and the underlying intent inference, based on different user sessions. Inspired by the reasoning capabilities of large language models (LLMs), we propose an adaptive in-context learning paradigm, which allows LLMs to draw tailored lessons from related sessions as demonstrations, enhancing the performance on target sessions. Specifically, we first employ retrieval augmented generation to identify nearest neighbor sessions, and then carefully design prompts to guide LLMs in executing both tasks on these neighbor sessions. To tackle reliability and hallucination challenges, we further introduce (1) a self-correction strategy promoting mutual improvements of the two tasks without supervision signals and (2) an auto-feedback mechanism for adaptive supervision based on the distinct mistakes made by LLMs on different neighbor sessions. Thereby, the target session can gain customized lessons for improved performance by observing the demonstrations of its neighbor sessions. Experiments on three real-world datasets demonstrate the effectiveness of our proposed method.

The goal of intent-aware session recommendation (ISR) approaches is to capture user intents within a session for accurate next-item prediction. However, the capability of these approaches is limited by assuming all sessions have a uniform and fixed number of intents. In reality, user sessions can vary, where the number of intentions may differ from one to another. Moreover, they can only learn user intents in the latent space, which further restricts the model's transparency. To ease these issues, we propose a simple yet effective paradigm for ISR motivated by the advanced reasoning capability of large language models (LLMs). Specifically, we first create an initial prompt to instruct LLMs to predict the next item by inferring varying user intents reflected in a session. Then, we propose an effective optimization mechanism to automatically optimize prompts with an iterative self-reflection. Finally, we leverage the robust generalizability of LLMs across diverse domains to efficiently select the optimal prompt for ISR. As such, the proposed paradigm effectively guides LLMs to identify varying user intents at a semantic level, thus delivering more accurate and comprehensible recommendations. Extensive experiments on three real-world datasets verify the superiority of our proposed method.

In this paper, we present Polarized message-passing (PMP), a novel paradigm to revolutionize the design of message-passing graph neural networks (GNNs). In contrast to existing methods, PMP captures the power of node-node similarity and dissimilarity to acquire dual sources of messages from neighbors. The messages are then coalesced to enable GNNs to learn expressive representations from sparse but strongly correlated neighbors. Three novel GNNs based on the PMP paradigm, namely PMP graph convolutional network (PMP-GCN), PMP graph attention network (PMP-GAT), and PMP graph PageRank network (PMP-GPN) are proposed to perform various downstream tasks. Theoretical analysis is also conducted to verify the high expressiveness of the proposed PMP-based GNNs. In addition, an empirical study of five learning tasks based on 12 real-world datasets is conducted to validate the performances of PMP-GCN, PMP-GAT, and PMP-GPN. The proposed PMP-GCN, PMP-GAT, and PMP-GPN outperform numerous strong message-passing GNNs across all five learning tasks, demonstrating the effectiveness of the proposed PMP paradigm.

Policy diversity, encompassing the variety of policies an agent can adopt, enhances reinforcement learning (RL) success by fostering more robust, adaptable, and innovative problem-solving in the environment. The environment in which standard RL operates is usually modeled with a Markov Decision Process (MDP) as the theoretical foundation. However, in many real-world scenarios, the rewards depend on an agent's history of states and actions leading to a non-MDP. Under the premise of policy diffusion initialization, non-MDPs may have unstructured expanding solution space due to varying historical information and temporal dependencies. This results in solutions having non-equivalent closed forms in non-MDPs. In this paper, deriving diverse solutions for non-MDPs requires policies to break through the boundaries of the current solution space through gradual dispersion. The goal is to expand the solution space, thereby obtaining more diverse policies. Specifically, we first model the sequences of states and actions by a transformer-based method to learn policy embeddings for dispersion in the solution space, since the transformer has advantages in handling sequential data and capturing long-range dependencies for non-MDP. Then, we stack the policy embeddings to construct a dispersion matrix as the policy diversity measure to induce the policy dispersion in the solution space and obtain a set of diverse policies. Finally, we prove that if the dispersion matrix is positive definite, the dispersed embeddings can effectively enlarge the disagreements across policies, yielding a diverse expression for the original policy embedding distribution. Experimental results of both non-MDP and MDP environments show that this dispersion scheme can obtain more expressive diverse policies via expanding the solution space, showing more robust performance than the recent learning baselines.

In an era of pervasive digitalization, the growing volume and variety of data streams poses a new challenge to the efficient running of data-driven optimization algorithms. Targeting scalable multiobjective evolution under large-instance data, this article proposes the general idea of using subsampled small-data tasks as helpful minions (i.e., auxiliary source tasks) to quickly optimize for large datasets-via an evolutionary multitasking framework. Within this framework, a novel computational resource allocation strategy is designed to enable the effective utilization of the minions while guarding against harmful negative transfers. To this end, an intertask empirical correlation measure is defined and approximated via Bayes' rule, which is then used to allocate resources online in proportion to the inferred degree of source-target correlation. In the experiments, the performance of the proposed algorithm is verified on: 1) sample average approximations of benchmark multiobjective optimization problems under uncertainty and 2) practical multiobjective hyperparameter tuning of deep neural network models. The results show that the proposed algorithm can obtain up to about 73% speedup relative to existing approaches, demonstrating its ability to efficiently tackle real-world multiobjective optimization involving evaluations on large datasets.

The multiple-choice knapsack problem (MCKP) is a classic NP-hard combinatorial optimization problem. Motivated by several significant real-world applications, this work investigates a novel variant of MCKP called the chance-constrained MCKP (CCMCKP), where item weights are random variables. In particular, we focus on the practical scenario of CCMCKP, in which the probability distributions of random weights are unknown and only sample data is available. We first present the problem formulation of CCMCKP and then establish the two benchmark sets. The first set contains synthetic instances, while the second set is designed to simulate a real-world application scenario of a telecommunication company. To solve CCMCKP, we propose a data-driven adaptive local search (DDALS) algorithm. Compared to existing stochastic optimization and distributionally robust optimization methods, the main novelty of DDALS lies in its data-driven solution evaluation approach, which does not make any assumptions about the underlying distributions and is highly effective even when faced with a high intensity of the chance constraint and a limited amount of sample data. Experimental results demonstrate the superiority of DDALS over the baselines on both the benchmarks. Finally, DDALS can serve as the baseline for future research, and the benchmark sets are open-sourced to further promote research on this challenging problem.

Recent developments in reinforcement learning (RL) have been able to derive optimal policies for sophisticated and capable agents, and shown to achieve human-level performance on a number of challenging tasks. Unfortunately, when it comes to multiagent systems (MASs), complexities, such as nonstationarity and partial observability bring new challenges to the field. Building a flexible and efficient multiagent RL (MARL) algorithm capable of handling complex tasks has to date remained an open challenge. This article presents a multiagent learning system with the evolution of social roles (eSRMA). The main interest is placed on solving the key issues in the definition and evolution of suitable roles, and optimizing the policies accompanied by social roles in MAS efficiently. Specifically, eSRMA incorporates and cultivates role division awareness of agents to improve the ability to deal with complex cooperative tasks. Each agent is assigned a role module, which can dynamically generate roles based on the individuals’ local observations. A novel MARL algorithm is designed as the principal driving force that governs the role-policy learning process by a role-attention credit assignment mechanism. Moreover, a role evolution process is developed to help agents dynamically choose appropriate roles in decision making. Comprehensive experiments on the StarCraft II micromanagement benchmarkhave demonstrated that eSRMA exhibits superiority in achieving higher learning capability and efficiency for multiple agents compared to the state-of-the-art MARL methods.

Transfer optimization enables data-efficient optimization of a target task by leveraging experiential priors from related source tasks. This is especially useful in multiobjective optimization settings where a set of tradeoff solutions is sought under tight evaluation budgets. In this article, we introduce a novel concept of inverse transfer in multiobjective optimization. Inverse transfer stands out by employing Bayesian inverse Gaussian process models to map performance vectors in the objective space to population search distributions in task-specific decision space, facilitating knowledge transfer through objective space unification . Building upon this idea, we introduce the first Inverse Transfer Evolutionary Multiobjective Optimizer (invTrEMO). A key highlight of invTrEMO is its ability to harness the common objective functions prevalent in many application areas, even when decision spaces do not precisely align between tasks. This allows invTrEMO to uniquely and effectively utilize information from heterogeneous source tasks as well. Furthermore, invTrEMO yields high-precision inverse models as a significant byproduct, enabling the generation of tailored solutions on-demand based on user preferences. Empirical studies on multi- and many-objective benchmark problems, as well as a practical case study, showcase the faster convergence rate and modeling accuracy of the invTrEMO relative to state-of-the-art evolutionary and Bayesian optimization algorithms. The source code of the invTrEMO is made available at https://github.com/LiuJ-2023/invTrEMO .

Automatic lung organ segmentation on computerized tomography images is crucial for lung disease diagnosis. However, the unlimited voxel values and class imbalance of lung organs can lead to false-negative/positive and leakage issues in numerous state-of-the-art methods. In addition, some lung organs are easily lost during therecycleddown/up-sample procedure, e.g., bronchioles and arterioles, which can cause severe discontinuity issue. Inspired by these, this article introduces an effective lung organ segmentation method called fuzzy attention-based border rendering feature orthogonal network, which 1) integrates an efficient transformer-like fuzzy-attention module into deep networks to cope with the uncertainty in feature representations; 2) decouples and depicts the lung organ regions as cube-trees by focusing only onrecycle-sampling border vulnerable points, rendering the severely discontinuous, false-negative/positive organ regions with two novel global-local cube-tree fusion and sparse patched feature orthogonal modules; 3) develops a multiscale self-knowledge guidance module to improve model performance and robustness. We have demonstrated the efficacy of proposed method on five challenging datasets of lung organ segmentation, i.e., airway and artery. All experimental results demonstrate that our method can achieve the favorable performance significantly.

Detecting and quantifying abnormal crowd motion emerging from complex interactions of individuals is paramount to ensure the safety of crowds. Crowd-level abnormal behaviors (CABs), e.g., counter flow and crowd turbulence, are proven to be the crucial causes of many crowd disasters. Unlike individual-level anomaly, CABs usually do not exhibit salient difference from the normal behaviors when observed locally and the scale of CABs could vary from one scenario to another. It is also challenging to quantify the risk level of these CABs from video surveillance. In this paper, we present an improved version of our crowd motion learning framework for CABs detection, multi-scale motion consistency network (MSMC-Net) with a dual-attention fusion process to accommodate both the spatio-temporal and scale variations of different CABs. In addition, we propose an assessment method to quantify the risk level of detected CABs based on the anomaly score generated from our MSMC-Net. The risk quantification is performed in an online and accumulated manner and it can reflect the risk level of CABs consistent with other offline assessment metrics (e.g., crowd pressure), but without the extraction of detailed crowd data (e.g., pedestrian trajectories). For empirical study, we evaluate our method on large-scale crowd event datasets, including UMN, Hajj and Love Parade. Experimental results show that MSMC-Net could improve the AUC performance by 7.9%, 12.2% and 29.5% on three datasets respectively, compared to the best results of the state-of-the-art methods.

Graph clusters (or communities) represent important graph structural information. In this paper, we presentDifferentiableClustering for graphATtention (DCAT). To the best of our knowledge, DCAT is the first solution that incorporates graph clustering into graph attention networks (GAT) to learn cluster-aware attention scores for semi-supervised learning tasks. In DCAT, we propose a novel approach to formunderlineating graph clustering as an auxiliary differentiable objective based on modunderlinearity maximization, which can be optimized together with the learning objective of GAT for a semi-supervised task. Specifically, we propose a solution to relaxing modunderlinearity maximization from a discrete optimization problem to a differentiable objective with theoretical guarantee so that we can learn cluster-aware attention scores by jointly learning from graph clustering and a semi-supervised learning task. To address the computational challenge, we further propose to reformunderlineate the constraint introduced by the clustering objective into a new form. Our analysis shows that DCAT allocates higher attention scores to nodes within the same cluster, allowing them to have a higher influence in node representation learning, and thus DCAT will generate better node representations for downstream applications. The experimental resunderlinets on commonly used datasets show that DCAT outperforms popunderlinear and state-of-the-art graph neural networks.

The pandemic of coronavirus disease 2019 (COVID-19) has led to a global public health crisis, which caused millions of deaths and billions of infections, greatly increasing the pressure on medical resources. With the continuous emergence of viral mutations, developing automated tools for COVID-19 diagnosis is highly desired to assist the clinical diagnosis and reduce the tedious workload of image interpretation. However, medical images in a single site are usually of a limited amount or weakly labeled, while integrating data scattered around different institutions to build effective models is not allowed due to data policy restrictions. In this article, we propose a novel privacy-preserving cross-site framework for COVID-19 diagnosis with multimodal data, seeking to effectively leverage heterogeneous data from multiple parties while preserving patients' privacy. Specifically, a Siamese branched network is introduced as the backbone to capture inherent relationships across heterogeneous samples. The redesigned network is capable of handling semisupervised inputs in multimodalities and conducting task-specific training, in order to improve the model performance of various scenarios. The framework achieves significant improvement compared with state-of-the-art methods, as we demonstrate through extensive simulations on real-world datasets.

Recent years have witnessed the great success of group buying (GB) in social e-commerce, opening up a new way of online shopping. In this business model, a user can launch a GB as an initiator to share her interested product with social friends. The GB is clinched once enough friends join in as participants to copurchase the shared product. As such, a successful GB depends on not only whether the initiator can find her interested product but also whether the friends are willing to join in as participants. Most existing recommenders are incompetent in such complex scenario, as they merely seek to help users find their preferred products and cannot help identify potential participants to join in a GB. To this end, we propose a novel joint product-participant recommendation (J2PRec) framework, which recommends both candidate products and participants for maximizing the success rate of a GB. Specifically, J2PRec first designs a relational graph embedding module, which effectively encodes the various relations in GB for learning enhanced user and product embeddings. It then jointly learns the product and participant recommendation tasks under a probabilistic framework to maximize the GB likelihood, i.e., boost the success rate of a GB. Extensive experiments on three real-world datasets demonstrate the superiority of J2PRec for GB recommendation.

Current one-stage methods for visual grounding encode the language query as one holistic sentence embedding before fusion with visual features for target localization. Such a formulation provides insufficient ability to model query at the word level, and therefore is prone to neglect words that may not be the most important ones for a sentence but are critical for the referred object. In this article, we propose Word2Pix: a one-stage visual grounding network based on the encoder-decoder transformer architecture that enables learning for textual to visual feature correspondence via word to pixel attention. Each word from the query sentence is given an equal opportunity when attending to visual pixels through multiple stacks of transformer decoder layers. In this way, the decoder can learn to model the language query and fuse language with the visual features for target prediction simultaneously. We conduct the experiments on RefCOCO, RefCOCO+, and RefCOCOg datasets, and the proposed Word2Pix outperforms the existing one-stage methods by a notable margin. The results obtained also show that Word2Pix surpasses the two-stage visual grounding models, while at the same time keeping the merits of the one-stage paradigm, namely, end-to-end training and fast inference speed. Code is available at https://github.com/azurerain7/Word2Pix.

Current study on next point-of-interest (POI) recommendation mainly explores user sequential transitions with the fine-grained individual-user POI check-in trajectories only, which suffers from the severe check-in data sparsity issue. In fact, coarse-grained signals (i.e., region- and global-level check-ins) in such sparse check-ins would also benefit to augment user preference learning. Specifically, our data analysis unveils that user movement exhibits noticeable patterns w.r.t. the regions of visited POIs. Meanwhile, the global all-user check-ins can help reflect sequential regularities shared by the crowd. We are, therefore, inspired to propose the MCMG: a Multi-Channel next POI recommendation framework with Multi-Granularity signals categorized from two orthogonal perspectives, i.e., fine-coarse grained check-ins at either POI/region level or local/global level. The MCMG is equipped with three modules, namely, global user behavior encoder, local multi-channel (i.e., region, category, and POI channels) encoder, and region-aware weighting strategy. Such design enables MCMG to be capable of capturing both fine- and coarse-grained sequential regularities as well as exploring the dynamic impact of multi-channel by differentiating the check-in patterns w.r.t. visited regions. Extensive experiments on four real-world datasets show that our MCMG significantly outperforms state-of-the-art next POI recommendation approaches.

Current session-based recommender systems (SBRSs) mainly focus on maximizing recommendation accuracy, while few studies have been devoted to improve diversity beyond accuracy. Meanwhile, it is unclear how the accuracy-oriented SBRSs perform in terms of diversity. In addition, the asserted “tradeoff” relationship between accuracy and diversity has been increasingly questioned in the literature. Toward the aforementioned issues, we conduct a holistic study to particularly examine the recommendation performance of representative SBRSs w.r.t. both accuracy and diversity, striving for better understanding of the diversity-related issues for SBRSs and providing guidance on designing diversified SBRSs. Particularly, for a fair and thorough comparison, we deliberately select state-of-the-art non-neural, deep neural, and diversified SBRSs by covering more scenarios with appropriate experimental setups, e.g., representative datasets, evaluation metrics, and hyper-parameter optimization technique. The source code can be obtained via github.com/qyin863/Understanding-Diversity-in-SBRSs . Our empirical results unveil that (1) non-diversified methods can also obtain satisfying performance on diversity, which can even surpass diversified ones, and (2) the relationship between accuracy and diversity is quite complex. Besides the “tradeoff” relationship, they can be positively correlated with each other, that is, having a same-trend (win–win or lose–lose) relationship, which varies across different methods and datasets. Additionally, we further identify three possible influential factors on diversity in SBRSs (i.e., granularity of item categorization, session diversity of datasets, and length of recommendation lists) and offer an intuitive guideline and a potential solution regarding learned item embeddings for more effective session-based recommendation.

Product bundling represents a prevalent marketing strategy in both offline stores and e-commerce systems. Despite its widespread use, previous studies on bundle recommendation face two significant limitations. Firstly, they rely on noisy datasets, where bundles are defined by heuristics, e.g., products co-purchased in the same session. Secondly, they target specific tasks by holding unrealistic assumptions, e.g., the availability of bundles for recommendation directly. This paper proposes to take a step back and considers the process of bundle recommendation from a holistic user experience perspective. We first construct high-quality bundle datasets with rich metadata, particularly bundle intents, through a carefully designed crowd-sourcing task. We then define a series of tasks that together, support all key steps in a typical bundle recommendation process, from bundle detection, completion and ranking, to explanation and auto-naming, whereby 19 research questions are raised correspondingly to guide the analysis. Finally, we conduct extensive experiments and analyses with representative recommendation models and large language models (LLMs), demonstrating the challenges and opportunities, especially with the emergence of LLMs. To summarize, our study contributes by introducing novel data sources, paving the way for new research avenues, and offering insights to guide product bundling in real e-commerce platforms.

Session-based recommendation (SR) has gained increasing attention in recent years. Quite a great amount of studies have been devoted to designing complex algorithms to improve recommendation performance, where deep learning methods account for the majority. However, most of these methods are black-box ones and ignore to provide moderate explanations to facilitate users’ understanding, which thus might lead to lowered user satisfaction and reduced system revenues. Therefore, in our study, we propose a generic Reinforced Explainable framework with Knowledge graph for Session-based recommendation (i.e., REKS), which strives to improve the existing black-box SR models (denoted as non-explainable ones) with Markov decision process. In particular, we construct a knowledge graph with session behaviors and treat SR models as part of the policy network of Markov decision process. Based on our particularly designed state vector, reward strategy, and loss function, the reinforcement learning (RL)-based framework not only achieves improved recommendation accuracy, but also provides appropriate explanations at the same time. Finally, we instantiate the REKS in five representative, state-of-the-art SR models (i.e., GRU4REC, NARM, SR-GNN, GCSAN, BERT4REC), whereby extensive experiments towards these methods on four datasets demonstrate the effectiveness of our framework on both recommendation and explanation tasks.

We present a novel loss formulation for efficient learning of complex dynamics from governing physics, typically described by partial differential equations (PDEs), using physics-informed neural networks (PINNs). In our experiments, existing versions of PINNs are seen to learn poorly in many problems, especially for complex geometries, as it becomes increasingly difficult to establish appropriate sampling strategy at the near boundary region. Overly dense sampling can adversely impede training convergence if the local gradient behaviors are too complex to be adequately modelled by PINNs. On the other hand, if the samples are too sparse, existing PINNs tend to overfit the near boundary region, leading to incorrect solution. To prevent such issues, we propose a new Boundary Connectivity (BCXN) loss function which provides linear local structure approximation (LSA) to the gradient behaviors at the boundary for PINN. Our BCXN-loss implicitly imposes local structure during training, thus facilitating fast physics-informed learning across entire prob-lem domains with order of magnitude sparser training samples. This LSA-PINN method shows a few orders of magnitude smaller errors than existing methods in terms of the standard L2-norm metric, while using dramatically fewer training samples and iterations. Our proposed LSA-PINN does not pose any requirement on the differentiable property of the networks, and we demonstrate its benefits and ease of implementation on both multi-layer perceptron and convolutional neural network versions as commonly used in current PINN literature.

Item consumption and social connection, as common user behaviors in many web applications, have been extensively studied. However, most current works separately perform either item consumption or social link prediction tasks, possibly with the help of the other as an auxiliary signal. Moreover, they merely consider the behaviors in a holistic manner yet neglect the multi-faceted motives behind them. For example, the intention of watching a movie could be killing time or watching it with friends; Likewise, one might connect with others due to friendships or colleagues. To fill this gap, we propose to Disentangle the multi-faceted Motives in each network (i.e., the user-item interaction network and social network) defined respectively by the two types of behaviors, for mutually-enhanced Joint Prediction (DMJP). Specifically, we first learn the disentangled user representations driven by motives of multi-facets in both networks. Thereafter, the mutual influence of the two networks is subtly discriminated at the facet-to-facet level. The fine-grained mutual influence is then exploited asymmetrically to help refine user representations in both networks, with the goal of achieving a mutually-enhanced joint item and social link prediction. Empirical studies on three public datasets showcase the superiority of DMJP over state-of-the-arts (SOTAs) on both tasks.

Conversational recommender systems (CRSs) have become crucial emerging research topics in the field of RSs, thanks to their natural advantages of explicitly acquiring user preferences via interactive conversations and revealing the reasons behind recommendations. However, the majority of current CRSs are text-based, which is less user-friendly and may pose challenges for certain users, such as those with visual impairments or limited writing and reading abilities. Therefore,for the first time, this paper investigates the potential of voice-based CRS (VCRSs) to revolutionize the way users interact with RSs in a natural, intuitive, convenient, and accessible fashion. To support such studies, we create two VCRSs benchmark datasets in the e-commerce and movie domains, after realizing the lack of such datasets through an exhaustive literature review. Specifically, we first empirically verify the benefits and necessity of creating such datasets. Thereafter, we convert the user-item interactions to text-based conversations through the ChatGPT-driven prompts for generating diverse and natural templates, and then synthesize the corresponding audios via the text-to-speech model. Meanwhile, a number of strategies are delicately designed to ensure the naturalness and high quality of voice conversations. On this basis, we further explore the potential solutions and point out possible directions to build end-to-end VCRSs by seamlessly extracting and integrating voice-based inputs, thus delivering performance-enhanced, self-explainable, and user-friendly VCRSs. Our study aims to establish the foundation and motivate further pioneering research in the emerging field of VCRSs. This aligns with the principles of explainable AI and AI for social good, viz., utilizing technology's potential to create a fair, sustainable, and just world. Our codes and datasets are available on GitHub (https://github.com/hyllll/VCRS ).

Privacy-preserving collaborative learning enables resource-constrained edge devices (e.g., Internet of Things (IoT) devices and smartphones) to build a knowledge-shared model while keeping individual data locally, achieving privacy preservation by designing an effective communication protocol. However, the learning paradigm raises high requirements for aligned input features of models, which is hard to realize in complicated IoT scenarios with various monitoring indicators. In this article, we propose a novel collaborative learning framework that is tolerant of IoT devices with unaligned feature spaces. Local bilevel optimizations for both model parameters and input features are performed iteratively in the training phase, in which the internal correlations of local sensor data provide additional guidance for the feature inference and completion. The scheme breaks hardware boundaries among various IoT devices in collaboration with the assistance of model inversion inference, which gains a new perspective on the utilization of model confidentiality and requires minimal modifications to the existing collaborative learning process. The framework achieves significant improvement compared with state-of-the-art methods, as we demonstrate through extensive simulations on real-world data sets.

Multi-party learning provides an effective approach for training a machine learning model, e.g., deep neural networks (DNNs), over decentralized data by leveraging multiple decentralized computing devices, subjected to legal and practical constraints. Different parties, so-called local participants, usually provide heterogenous data in a decentralized mode, leading to non-IID data distributions across different local participants which pose a notorious challenge for multi-party learning. To address this challenge, we propose a novel heterogeneous differentiable sampling (HDS) framework. Inspired by the dropout strategy in DNNs, a data-driven network sampling strategy is devised in the HDS framework, with differentiable sampling rates which allow each local participant to extract from a common global model the optimal local model that best adapts to its own data properties so that the size of the local model can be significantly reduced to enable more efficient inference. Meanwhile, co-adaptation of the global model via learning such local models allows for achieving better learning performance under non-IID data distributions and speeds up the convergence of the global model. Experiments have demonstrated the superiority of the proposed method over several popular multi-party learning techniques in the multi-party settings with non-IID data distributions.

Recently, one critical issue looms large in the field of recommender systems - there are no effective benchmarks for rigorous evaluation - which consequently leads to unreproducible evaluation and unfair comparison. We, therefore, conduct studies from the perspectives of practical theory and experiments, aiming at benchmarking recommendation for rigorous evaluation. Regarding the theoretical study, a series of hyper-factors affecting recommendation performance throughout the whole evaluation chain are systematically summarized and analyzed via an exhaustive review on 141 papers published at eight top-tier conferences within 2017-2020. We then classify them into model-independent and model-dependent hyper-factors, and different modes of rigorous evaluation are defined and discussed in-depth accordingly. For the experimental study, we release DaisyRec 2.0 library by integrating these hyper-factors to perform rigorous evaluation, whereby a holistic empirical study is conducted to unveil the impacts of different hyper-factors on recommendation performance. Supported by the theoretical and experimental studies, we finally create benchmarks for rigorous evaluation by proposing standardized procedures and providing performance of ten state-of-the-arts across six evaluation metrics on six datasets as a reference for later study. Overall, our work sheds light on the issues in recommendation evaluation, provides potential solutions for rigorous evaluation, and lays foundation for further investigation.

Transfer regression is a practical and challenging problem with important applications in various domains, such as engineering design and localization. Capturing the relatedness of different domains is the key of adaptive knowledge transfer. In this paper, we investigate an effective way of explicitly modelling domain relatedness through transfer kernel, a transfer-specified kernel that considers domain information in the covariance calculation. Specifically, we first give the formal definition of transfer kernel, and introduce three basic general forms that well cover existing related works. To cope with the limitations of the basic forms in handling complex real-world data, we further propose two advanced forms. Corresponding instantiations of the two forms are developed, namely${Trk}_{\alpha \beta }$and${Trk}_{\omega }$based on multiple kernel learning and neural networks, respectively. For each instantiation, we present a condition with which the positive semi-definiteness is guaranteed and a semantic meaning is interpreted to the learned domain relatedness. Moreover, the condition can be easily used in the learning ofTrGP$_{\alpha \beta }$andTrGP$_{\omega }$that are the Gaussian process models with the transfer kernels${Trk}_{\alpha \beta }$and${Trk}_{\omega }$respectively. Extensive empirical studies show the effectiveness ofTrGP$_{\alpha \beta }$andTrGP$_{\omega }$on domain relatedness modelling and transfer adaptiveness.

Multi-source transfer regression is a practical and challenging problem where capturing the diverse relatedness of different domains is the key of adaptive knowledge transfer. In this paper, we propose an effective way of explicitly modeling the domain relatedness of each domain pair through transfer kernel learning. Specifically, we first discuss the advantages and disadvantages of existing transfer kernels in handling the multi-source transfer regression problem. To cope with the limitations of the existing transfer kernels, we further propose a novel multi-source transfer kernel$k_{ms}$. The proposed$k_{ms}$assigns a learnable parametric coefficient to model the relatedness of each inter-domain pair, and simultaneously regulates the relatedness of the intra-domain pair to be 1. Moreover, to capture the heterogeneous data characteristics of multiple domains,$k_{ms}$exploits different standard kernels for different domain pairs. We further provide a theorem that not only guarantees the positive semi-definiteness of$k_{ms}$but also conveys a semantic interpretation to the learned domain relatedness. Moreover, the theorem can be easily used in the learning of the corresponding transfer Gaussian process model with$k_{ms}$. Extensive empirical studies show the effectiveness of our proposed method on domain relatedness modelling and transfer performance.

The performance of machine learning algorithms heavily relies on the availability of a large amount of training data. However, in reality, data usually reside in distributed parties such as different institutions and may not be directly gathered and integrated due to various data policy constraints. As a result, some parties may suffer from insufficient data available for training machine learning models. In this article, we propose a multiparty dual learning (MPDL) framework to alleviate the problem of limited data with poor quality in an isolated party. Since the knowledge-sharing processes for multiple parties always emerge in dual forms, we show that dual learning is naturally suitable to handle the challenge of missing data, and explicitly exploits the probabilistic correlation and structural relationship between dual tasks to regularize the training process. We introduce a feature-oriented differential privacy with mathematical proof, in order to avoid possible privacy leakage of raw features in the dual inference process. The approach requires minimal modifications to the existing multiparty learning structure, and each party can build flexible and powerful models separately, whose accuracy is no less than nondistributed self-learning approaches. The MPDL framework achieves significant improvement compared with state-of-the-art multiparty learning methods, as we demonstrated through simulations on real-world datasets.

Many real-world problems, such as airfoil design, involve optimizing a black-box expensive objective function over complex-structured input space (e.g., discrete space or non-Euclidean space). By mapping the complex-structured input space into a latent space of dozens of variables, a two-stage procedure labeled as generative model-based optimization (GMO), in this article, shows promise in solving such problems. However, the latent dimension of GMO is hard to determine, which may trigger the conflicting issue between desirable solution accuracy and convergence rate. To address the above issue, we propose a multiform GMO approach, namely, generative multiform optimization (GMFoO), which conducts optimization over multiple latent spaces simultaneously to complement each other. More specifically, we devise a generative model which promotes a positive correlation between latent spaces to facilitate effective knowledge transfer in GMFoO. And furthermore, by using Bayesian optimization (BO) as the optimizer, we propose two strategies to exchange information between these latent spaces continuously. Experimental results are presented on airfoil and corbel design problems and an area maximization problem as well to demonstrate that our proposed GMFoO converges to better designs on a limited computational budget.

Collision-avoidance control for UAV swarm has recently drawn great attention due to its significant implications in many industrial and commercial applications. However, traditional collision-avoidance models for UAV swarm tend to focus on avoidance at individual UAV level, and no explicit strategy is designed for avoidance among multiple UAV groups. When directly applying these models for multigroup UAV scenarios, the deadlock situation may happen. A group of UAVs may be temporally blocked by other groups in a narrow space and cannot progress toward achieving its goal. To this end, this article proposes a modeling and optimization approach to multigroup UAV collision avoidance. Specifically, group level collision detection and adaption mechanism are introduced, efficiently detecting potential collisions among different UAV groups and restructuring a group into subgroups for better collision and deadlock avoidance. A two-level control model is then designed for realizing collision avoidance among UAV groups and of UAVs within each group. Finally, an evolutionary multitask optimization method is introduced to effectively calibrate the parameters that exist in different levels of our control model, and an adaptive fitness evaluation strategy is proposed to reduce computation overhead in simulation-based optimization. The simulation results show that our model has superior performances in deadlock resolution, motion stability, and distance maintenance in multigroup UAV scenarios compared to the state-of-the-art collision-avoidance models. The model optimization results also show that our model optimization method can largely reduce execution time for computationally-intensive optimization process that involves UAV swarm simulation.

In today's digital world, we are faced with an explosion of data and models produced and manipulated by numerous large-scale cloud-based applications. Under such settings, existing transfer evolutionary optimization (TrEO) frameworks grapple with simultaneously satisfying two important quality attributes, namely: 1) scalability against a growing number of source tasks and 2) online learning agility against sparsity of relevant sources to the target task of interest. Satisfying these attributes shall facilitate practical deployment of transfer optimization to scenarios with big task instances, while curbing the threat of negative transfer. While applications of existing algorithms are limited to tens of source tasks, in this article, we take a quantum leap forward in enabling more than two orders of magnitude scale-up in the number of tasks; that is, we efficiently handle scenarios beyond 1000 source task instances. We devise a novel TrEO framework comprising two co-evolving species for joint evolutions in the space of source knowledge and in the search space of solutions to the target problem. In particular, co-evolution enables the learned knowledge to be orchestrated on the fly, expediting convergence in the target optimization task. We have conducted an extensive series of experiments across a set of practically motivated discrete and continuous optimization examples comprising a large number of source task instances, of which only a small fraction indicate source-target relatedness. The experimental results show that not only does our proposed framework scale efficiently with a growing number of source tasks but is also effective in capturing relevant knowledge against sparsity of related sources, fulfilling the two salient features of scalability and online learning agility.

In dealing with the expensive multiobjective optimization problem, some algorithms convert it into a number of single-objective subproblems for optimization. At each iteration, these algorithms conduct surrogate-assisted optimization on one or multiple subproblems. However, these subproblems may be unnecessary or resolved. Operating on such subproblems can cause server inefficiencies, especially in the case of expensive optimization. To overcome this shortcoming, we propose an adaptive subproblem selection (ASS) strategy to identify the most promising subproblems for further modeling. To better leverage the cross information between the subproblems, we use the collaborative multioutput Gaussian process surrogate to model them jointly. Moreover, the commonly used acquisition functions (also known as infill criteria) are investigated in this article. Our analysis reveals that these acquisition functions may cause severe imbalances between exploitation and exploration in multiobjective optimization scenarios. Consequently, we develop a new acquisition function, namely, adaptive lower confidence bound (ALCB), to cope with it. The experimental results on three different sets of benchmark problems indicate that our proposed algorithm is competitive. Beyond that, we also quantitatively validate the effectiveness of the ASS strategy, the CoMOGP model, and the ALCB acquisition function.

Graph classification aims to predict the label associated with a graph and is an important graph analytic task with widespread applications. Recently, graph neural networks (GNNs) have achieved state-of-the-art results on purely supervised graph classification by virtue of the powerful representation ability of neural networks. However, almost all of them ignore the fact that graph classification usually lacks reasonably sufficient labeled data in practical scenarios due to the inherent labeling difficulty caused by the high complexity of graph data. The existing semisupervised GNNs typically focus on the task of node classification and are incapable to deal with graph classification. To tackle the challenging but practically useful scenario, we propose a novel and general semisupervised GNN framework for graph classification, which takes full advantage of a slight amount of labeled graphs and abundant unlabeled graph data. In our framework, we train two GNNs as complementary views for collaboratively learning high-quality classifiers using both labeled and unlabeled graphs. To further exploit the view itself, we constantly select pseudo-labeled graph examples with high confidence from its own view for enlarging the labeled graph dataset and enhancing predictions on graphs. Furthermore, the proposed framework is investigated on two specific implementation regimes with a few labeled graphs and the extremely few labeled graphs, respectively. Extensive experimental results demonstrate the effectiveness of our proposed semisupervised GNN framework for graph classification on several benchmark datasets.

Reading psychology believes text comprehension to involve a complex psychological construction process, with the reader mind being a dynamic associative system that stores an abundance of schemata. For Chinese text, in particular, the unique ideographic writing system allows its lansign to trigger semantic association and schema recalling without the need of phonetics. In contrast to previous research efforts on text classification problems, in this paper we present an interdisciplinary modeling approach that draws inspirations from the cognitive principles of ideography, schema theory and deep learning to study Chinese text classification. Specifically, we first propose a Radical-guided Associative Model (RAM) for preliminary cognitive imitation, which comprises two coupled spaces, namely the Literal Space and Associative Space. Then, taking consideration of the schemata acquired from the mind of a reader which plays a important role in influencing text-dependent information revision, we extend RAM with a systematic Schema-aware Radical-guided Associative Model (SRAM) that embeds label semantics as essential text-independent human knowledge for real-world abstraction. In SRAM, the Schema Space is introduced and a Schema Attention module is proposed with a novel loss paradigm that includes the linkage and interaction between text-dependent prior concepts and text-independent label schemata. Extensive experiments on three real-world datasets demonstrate the effectiveness and rationality of our proposed method.

Neural architecture search (NAS) has attracted much attention in recent years. It automates the neural network construction for different tasks, which is traditionally addressed manually. In the literature, evolutionary optimization (EO) has been proposed for NAS due to its strong global search capability. However, despite the success enjoyed by EO, it is worth noting that existing EO algorithms for NAS are often very computationally expensive, which makes these algorithms unpractical in reality. Keeping this in mind, in this article, we propose an efficient memetic algorithm (MA) for automated convolutional neural network (CNN) architecture search. In contrast to existing EO algorithms for CNN architecture design, a new cell-based architecture search space, and new global and local search operators are proposed for CNN architecture search. To further improve the efficiency of our proposed algorithm, we develop a one-epoch-based performance estimation strategy without any pretrained models to evaluate each found architecture on the training datasets. To investigate the performance of the proposed method, comprehensive empirical studies are conducted against 34 state-of-the-art peer algorithms, including manual algorithms, reinforcement learning (RL) algorithms, gradient-based algorithms, and evolutionary algorithms (EAs), on widely used CIFAR10 and CIFAR100 datasets. The obtained results confirmed the efficacy of the proposed approach for automated CNN architecture design.

Deep reinforcement learning (DRL) policies have been shown to be deceived by perturbations (e.g., random noise or intensional adversarial attacks) on state observations that appear at test time but are unknown during training. To increase the robustness of DRL policies, previous approaches assume that explicit adversarial information can be added into the training process, to achieve generalization ability on these perturbed observations as well. However, such approaches not only make robustness improvement more expensive but may also leave a model prone to other kinds of attacks in the wild. In contrast, we propose an adversary agnostic robust DRL paradigm that does not require learning from predefined adversaries. To this end, we first theoretically show that robustness could indeed be achieved independently of the adversaries based on a policy distillation (PD) setting. Motivated by this finding, we propose a new PD loss with two terms: 1) a prescription gap maximization (PGM) loss aiming to simultaneously maximize the likelihood of the action selected by the teacher policy and the entropy over the remaining actions and 2) a corresponding Jacobian regularization (JR) loss that minimizes the magnitude of gradients with respect to the input state. The theoretical analysis substantiates that our distillation loss guarantees to increase the prescription gap and hence improves the adversarial robustness. Furthermore, experiments on five Atari games firmly verify the superiority of our approach compared to the state-of-the-art baselines.

Multitask Gaussian process (MTGP) is a well-known nonparametric Bayesian model for learning correlated tasks effectively by transferring knowledge across tasks. But current MTGPs are usually limited to the multitask scenario defined in the same input domain, leaving no space for tackling the heterogeneous case, i.e., the features of input domains vary over tasks. To this end, this article presents a novel heterogeneous stochastic variational linear model of coregionalization (HSVLMC) model for simultaneously learning the tasks with varied input domains. Particularly, we develop the stochastic variational framework with Bayesian calibration that: 1) infers posterior domain mappings to consider the effect of dimensionality reduction raised by domain mappings for achieving effective input alignment and 2) employs a residual modeling strategy to leverage the inductive bias brought by prior domain mappings for better-model inference. Finally, the superiority of the proposed model against existing heterogeneous LMC models has been extensively verified on diverse heterogeneous multitask cases and a practical multifidelity steam turbine exhaust case.

Transfer evolutionary optimization (TrEO) has emerged as a computational paradigm to leverage related problem-solving information from various source tasks to boost convergence rates in a target task. State-of-the-art Tr EO algorithms have utilized a source-target similarity capture method with probabilistic priors that grants the ability to reduce negative transfers. A recent work makes use of an additional solution representation learning module to induce high ordinal correlation between source and target objective functions through source-to-target search space mappings, with the aim of promoting positive transfers between them. However, current implementations of this approach are found to be data-intensive - calling for all generated source data to be cached - leading to high storage costs in practice. As an alternative, this paper investigates the feasibility of a data-lean variant of the aforesaid approach, labeled as (1, G)-TrEO, in which only the first and final (Gth) generations of source data are used for solution representation learning and transfer. We conduct experimental analyses of (1, G)-TrEO using multi-objective benchmark functions as well as a practical example in vehicle crashworthiness design. Our results show that a simple data-lean transfer optimizer is able to achieve competitive performance. While this paper presents a first investigation of (1, G)-TrEO, we hope that the findings would inspire future forms of data-lean TrEO algorithms.

Evolutionary multitasking optimization (EMTO) solves multiple optimization tasks meanwhile in the framework of evolutionary algorithm, aiming at improving the solving performance on each task via knowledge transfer among tasks. As one of the representative EMTO algorithms, multifactorial evolutionary algorithm (MFEA) has attracted great attention and has been used to solve many optimization problems. However, most of MFEAs tend to suffer from premature convergence. To deal with this issue, this article designs a novel MFEA by integrating differential evolution, and a hybrid of global and local search strategies, named MFDE-GLS for short. Particularly, the global search strategy is based on an opposition-based learning and a Gaussian perturbation to improve the search ability and maintain population diversity. A local search strategy is introduced by combining 1-dimension search and n-dimension search to accelerate the convergence. Moreover, a new environmental selection mechanism is developed to keep the elite individuals while maintaining the population diversity based on the affinity propagation clustering method. Comprehensive experiments were conducted on both single-objective and multi-objective multi-task benchmark problems to show the effectiveness of the proposed algorithm.

Model robustness is vital for the reliable deployment of machine learning models in real-world applications. Recent studies have shown that data augmentation can result in model over-relying on features in the low-frequency domain, sacrificing performance against low-frequency corruptions, highlighting a connection between frequency and robustness. Here, we take one step further to more directly study the frequency bias of a model through the lens of its Jacobians and its implication to model robustness. To achieve this, we propose Jacobian frequency regularization for models' Jacobians to have a larger ratio of low-frequency components. Through experiments on four image datasets, we show that biasing classifiers towards low (high)-frequency components can bring performance gain against high (low)-frequency corruption and adversarial perturbation, albeit with a tradeoff in performance for low (high)-frequency corruption. Our approach elucidates a more direct connection between the frequency bias and robustness of deep learning models.

Existing studies on next point-of-interest (POI) recommendation mainly attempt to learn user preference from the past and current sequential behaviors. They, however, completely ignore the impact of future behaviors on the decision-making, thus hindering the quality of user preference learning. Intuitively, users' next POI visits may also be affected by their multi-step future behaviors, as users may often have activity planning in mind. To fill this gap, we propose a novel Context-aware Future Preference inference Recommender (CFPRec) to help infer user future preference in a self-ensembling manner. In particular, it delicately derives multi-step future preferences from the learned past preference thanks to the periodic property of users' daily check-ins, so as to implicitly mimic user’s activity planning before her next visit. The inferred future preferences are then seamlessly integrated with the current preference for more expressive user preference learning. Extensive experiments on three datasets demonstrate the superiority of CFPRec against state-of-the-arts.

Forensic systems have recently been studied to detect and prevent deepfakes abuses such as fake personas, frauds, misinformation, or harassment. At the same time, anti-forensic deepfakes are being investigated to understand the gaps in these detection systems and pave the way for improvement. In this paper, we investigate the threat of anti-forensic fake personas, where a fraudster creates a fake personal profile from multiple anti-forensic deepfake images portraying a single identity. To comprehensively study this threat model, we consider three approaches that an attacker may use to conduct such attacks, encompassing both white- and black-box scenarios. A range of defense strategies is then proposed with the aim to improve the robustness of current forensic systems against such threats. Experimental result shows that while the attacks can bypass current detection, our proposed defense approaches that consider the multi-image nature of a fake persona can effectively mitigate this threat by lowering the attack success rate.

Mel frequency cepstral coefficient (MFCC) features are widely used in applications such as keyword spotting, bearing fault detection and heart sound classification. This work proposes a low power MFCC engine that enables its use for battery-powered edge applications. Three hardware algorithm co-optimizations were adopted to achieve energy efficient MFCC hardware implementation. The approximated MFCC features due to the optimizations still allows good accuracy when deployed in several applications such as keyword spotting and bearing fault detection, reporting negligible accuracy drop of $\le 1.5$%. The proposed MFCC hardware consumes only 128nW at 0.3V supply and occupies only 0.08mm2in 40nm CMOS technology, which are $5 \times $ and $2.75 \times $ power and area reduction respectively when compared to the prior arts.

Resistive random access memory (RRAM) based computing-in-memory (CIM) is attractive for edge artificial intelligence (AI) applications, thanks to its excellent energy efficiency, compactness and high parallelism in matrix vector multiplication (MatVec) operations. However, existing RRAM-based CIM designs often require complex programming scheme to precisely control the RRAM cells to reach the desired resistance states so that the neural network classification accuracy is maintained. This leads to large area and energy overhead as well as low RRAM area utilization. Hence, compact RRAM-based CIM with simple pulse-based programming scheme is thus more desirable. To achieve this, we propose a chip-in-the-loop training approach to compensate for the network performance drop due to the stochastic behavior of the RRAM cells. Note that, although the target RRAM cell here is a two-state RRAM (i.e binary, having only high and low resistance states), their inherent analog resistance values are used in the CIM operation. Our experiment using a 4-layer fully-connected binary neural network (BNN) showed that after retraining, the RRAM-based network accuracy can be recovered, regardless of the RRAM resistance distribution and $\frac{\text{R}_{\text{HRS}}}{\text{R}_{\text{LRS}}}$ resistance ratio.

Policy distillation (PD) has been widely studied in deep reinforcement learning (RL), while existing PD approaches assume that the demonstration data (i.e., state-action pairs in frames) in a decision making sequence is uniformly distributed. This may bring in unwanted bias since RL is a reward maximizing process instead of simple label matching. Given such an issue, we denote the frame importance as its contribution to the expected reward on a particular frame, and hypothesize that adapting such frame importance could benefit the performance of the distilled student policy. To verify our hypothesis, we analyze why and how frame importance matters in RL settings. Based on the analysis, we propose an importance prioritized PD framework that highlights the training on important frames, so as to learn efficiently. Particularly, the frame importance is measured by the reciprocal of weighted Shannon entropy from a teacher policy's action prescriptions. Experiments on Atari games and policy compression tasks show that capturing the frame importance significantly boosts the performance of the distilled policies.

Adversarial examples in automatic speech recognition (ASR) are naturally sounded by humans yet capable of fooling well trained ASR models to transcribe incorrectly. Existing audio adversarial examples are typically constructed by adding constrained perturbations on benign audio inputs. Such attacks are therefore generated with an audio dependent assumption. For the first time, we propose the Speech Synthesising based Attack (SSA), a novel threat model that constructs audio adversarial examples entirely from scratch, i.e., without depending on any existing audio to fool cutting-edge ASR models. To this end, we introduce a conditional variational auto-encoder (CVAE) as the speech synthesiser. Meanwhile, an adaptive sign gradient descent algorithm is proposed to solve the adversarial audio synthesis task. Experiments on three datasets (i.e., Audio Mnist, Common Voice, and Librispeech) show that our method could synthesise naturally sounded audio adversarial examples to mislead the start-of-the-art ASR models. Our web-page containing generated audio demos is at https://sites.google.com/view/ssa-asr/home.

Recently, neural implicit surfaces learning by volume rendering has become popular for multi-view reconstruction. However, one key challenge remains: existing approaches lack explicit multi-view geometry constraints, hence usually fail to generate geometry-consistent surface reconstruction. To address this challenge, we propose geometry-consistent neural implicit surfaces learning for multi-view reconstruction. We theoretically analyze that there exists a gap between the volume rendering integral and point-based signed distance function (SDF) modeling. To bridge this gap, we directly locate the zero-level set of SDF networks and explicitly perform multi-view geometry optimization by leveraging the sparse geometry from structure from motion (SFM) and photometric consistency in multi-view stereo. This makes our SDF optimization unbiased and allows the multi-view geometry constraints to focus on the true surface optimization. Extensive experiments show that our proposed method achieves high-quality surface reconstruction in both complex thin structures and large smooth regions, thus outperforming the state-of-the-arts by a large margin.

Product bundling is a commonly-used marketing strategy in both offline retailers and online e-commerce systems. Current research on bundle recommendation is limited by: (1) noisy datasets, where bundles are defined by heuristics, e.g., products co-purchased in the same session; and (2) specific tasks, holding unrealistic assumptions, e.g., the availability of bundles for recommendation directly. In this paper, we propose to take a step back and consider the process of bundle recommendation from a holistic user experience perspective. We first construct high-quality bundle datasets with rich meta information, particularly bundle intents, through a carefully designed crowd-sourcing task. We then define a series of tasks that together, support all key steps in a typical bundle recommendation process, from bundle detection, completion, ranking, to explanation and auto-naming. Finally, we conduct extensive experiments and in-depth analysis that demonstrate the challenges of bundle recommendation, arising from the need for capturing complex relations among users, products, and bundles, as well as the research opportunities, especially in graph-based neural methods. To sum up, our study delivers new data sources, opens up new research directions, and provides useful guidance for product bundling in real e-commerce platforms. Our datasets are available at GitHub (\urlhttps://github.com/BundleRec/bundle_recommendation ).

Biological brains have a natural capacity for resolving certain classification tasks. Studies on biologically plausible spiking neurons, architectures and mechanisms of artificial neural systems that closely match biological observations while giving high classification performance are gaining momentum. Spiking neural P systems (SN P systems) are a class of membrane computing models and third-generation neural networks that are based on the behavior of biological neural cells and have been used in various engineering applications. Furthermore, SN P systems are characterized by a highly flexible structure that enables the design of a machine learning algorithm by mimicking the structure and behavior of biological cells without the over-simplification present in neural networks. Based on this aspect, this paper proposes a novel type of SN P system, namely, layered SN P system (LSN P system), to solve classification problems by supervised learning. The proposed LSN P system consists of a multi-layer network containing multiple weighted fuzzy SN P systems with adaptive weight adjustment rules. The proposed system employs specific ascending dimension techniques and a selection method of output neurons for classification problems. The experimental results obtained using benchmark datasets from the UCI machine learning repository and MNIST dataset demonstrated the feasibility and effectiveness of the proposed LSN P system. More importantly, the proposed LSN P system presents the first SN P system that demonstrates sufficient performance for use in addressing real-world classification problems.

Detecting anomalies in videos is a fundamental issue in public security. The majority of existing deep learning methods often perform anomaly detection based on the behavior or the trajectory of a single target. However, due to the overlaps of the crowd and the low-resolution of monitoring images, the segmentation of population is hard to implement and the features cannot be learned thoroughly, which make the methods be easily disturbed by visual elements and thus may lead to false detection sometimes. To tackle these problems, we propose the influence-aware attention to learn the representative attributes of the whole crowd. Walking pedestrians can be divided into numbers of flows, and in this paper, we aim to measure the consistency of movement patterns in the same stream and the interactions between different streams. Meanwhile, great importance is given to the relation between pedestrians and the circumstance for certain anomalies occur as a result of environmental issues. Specifically, the influence-aware attention module is composed of the motion attention and the location attention, which is designed to quantify the relations in the scene from spatial and temporal aspects. For the lack of abnormal samples, we utilize a dual generator-based framework to learn interactions among normal scenes. Experimental results on six benchmarks verify the effectiveness and robustness of our proposed method.

Evolutionary multitasking, which solves multiple optimization tasks simultaneously, has gained increasing research attention in recent years. By utilizing the useful information from related tasks while solving the tasks concurrently, improved performance has been shown in various problems. Despite the success enjoyed by the existing evolutionary multitasking algorithms, still there is a lack of theoretical studies guaranteeing faster convergence compared to the conventional single task case. To analyze the effects of transferred information from related tasks, in this article, we first put forward a novel multitask gradient descent (MTGD) algorithm, which enhances the standard gradient descent updates with a multitask interaction term. The convergence of the resulting MTGD is derived. Furthermore, we present the first proof of faster convergence of MTGD relative to its single task counterpart. Utilizing MTGD, we formulate a gradient-free evolutionary multitasking algorithm called multitask evolution strategies (MTESs). Importantly, the single task evolution strategies (ESs) we utilize are shown to asymptotically approximate gradient descent and, hence, the faster convergence results derived for MTGD extend to the case of MTES as well. Numerical experiments comparing MTES with single task ES on synthetic benchmarks and practical optimization examples serve to substantiate our theoretical claim.

Dynamic multiobjective optimization problem (DMOP) denotes the multiobjective optimization problem, which contains objectives that may vary over time. Due to the widespread applications of DMOP existed in reality, DMOP has attracted much research attention in the last decade. In this article, we propose to solve DMOPs via an autoencoding evolutionary search. In particular, for tracking the dynamic changes of a given DMOP, an autoencoder is derived to predict the moving of the Pareto-optimal solutions based on the nondominated solutions obtained before the dynamic occurs. This autoencoder can be easily integrated into the existing multiobjective evolutionary algorithms (EAs), for example, NSGA-II, MOEA/D, etc., for solving DMOP. In contrast to the existing approaches, the proposed prediction method holds a closed-form solution, which thus will not bring much computational burden in the iterative evolutionary search process. Furthermore, the proposed prediction of dynamic change is automatically learned from the nondominated solutions found along the dynamic optimization process, which could provide more accurate Pareto-optimal solution prediction. To investigate the performance of the proposed autoencoding evolutionary search for solving DMOP, comprehensive empirical studies have been conducted by comparing three state-of-the-art prediction-based dynamic multiobjective EAs. The results obtained on the commonly used DMOP benchmarks confirmed the efficacy of the proposed method.

Bayesian optimization (BO) is well known to be sample efficient for solving black-box problems. However, BO algorithms may get stuck in suboptimal solutions even with plenty of samples. Intrinsically, such a suboptimal problem of BO can attribute to the poor surrogate accuracy of the trained Gaussian process (GP), particularly that in the regions where the optimal solutions locate. Hence, we propose to build multiple GP models instead of a single GP surrogate to complement each other, thus resolving the suboptimal problem of BO. Nevertheless, according to the bias-variance tradeoff equation, the individual prediction errors can increase when increasing the diversity of models, which may lead even worse overall surrogate accuracy. On the other hand, based on the theory of the Rademacher complexity, it has been proven that exploiting the agreement of models on unlabeled information can reduce the complexity of hypothesis space, therefore achieving the required surrogate accuracy with fewer samples. Such value of model agreement has been extensively demonstrated for co-training style algorithms to boost model accuracy with a small portion of samples. Inspired by the above, we propose a novel BO algorithm labeled as co-learning BO (CLBO), which exploits both model diversity and agreement on unlabeled information to improve the overall surrogate accuracy with limited samples, therefore achieving more efficient global optimization. Through tests on five numerical toy problems and three engineering benchmarks, the effectiveness of the proposed CLBO has been well demonstrated.

In this article, we present a novel matrix-factorization-based model, labeled here as Vicinal vertex allocated matrix factorization (VVAMo), for uncovering clusters in network data. Different from the past related efforts of network clustering, which consider the edge structure, vertex features, or both in their design, the proposed model includes the additional detail on vertex inclinations with respect to topology and features into the learning. In particular, by taking the latent preferences between vicinal vertices into consideration, VVAMo is then able to uncover network clusters composed of proximal vertices that share analogous inclinations, and correspondingly high structural and feature correlations. To ensure such clusters are effectively uncovered, we propose a unified likelihood function for VVAMo and derive an alternating algorithm for optimizing the proposed function. Subsequently, we provide the theoretical analysis of VVAMo, including the convergence proof and computational complexity analysis. To investigate the effectiveness of the proposed model, a comprehensive empirical study of VVAMo is conducted using extensive commonly used realistic network datasets. The results obtained show that VVAMo attained superior performances over existing classical and state-of-the-art approaches.

Deep kernel learning (DKL) leverages the connection between the Gaussian process (GP) and neural networks (NNs) to build an end-to-end hybrid model. It combines the capability of NN to learn rich representations under massive data and the nonparametric property of GP to achieve automatic regularization that incorporates a tradeoff between model fit and model complexity. However, the deterministic NN encoder may weaken the model regularization of the following GP part, especially on small datasets, due to the free latent representation. We, therefore, present a complete deep latent-variable kernel learning (DLVKL) model wherein the latent variables perform stochastic encoding for regularized representation. We further enhance the DLVKL from two aspects: 1) the expressive variational posterior through neural stochastic differential equation (NSDE) to improve the approximation quality and 2) the hybrid prior taking knowledge from both the SDE prior and the posterior to arrive at a flexible tradeoff. Extensive experiments imply that DLVKL-NSDE performs similar to the well-calibrated GP on small datasets, and shows superiority on large datasets.

Gaussian process classification (GPC) provides a flexible and powerful statistical framework describing joint distributions over function space. Conventional GPCs, however, suffer from: 1) poor scalability for big data due to the full kernel matrix and 2) intractable inference due to the non-Gaussian likelihoods. Hence, various scalable GPCs have been proposed through: 1) the sparse approximation built upon a small inducing set to reduce the time complexity and 2) the approximate inference to derive analytical evidence lower bound (ELBO). However, these scalable GPCs equipped with analytical ELBO are limited to specific likelihoods or additional assumptions. In this work, we present a unifying framework that accommodates scalable GPCs using various likelihoods. Analogous to GP regression (GPR), we introduce additive noises to augment the probability space for: 1) the GPCs with step, (multinomial) probit, and logit likelihoods via the internal variables and 2) particularly, the GPC using softmax likelihood via the noise variables themselves. This leads to unified scalable GPCs with analytical ELBO by using variational inference. Empirically, our GPCs showcase superiority on extensive binary/multiclass classification tasks with up to two million data points.

Adversarial attack can be deemed as a necessary prerequisite evaluation procedure before the deployment of any reinforcement learning (RL) policy. Most existing approaches for generating adversarial attacks are gradient based and are extensive, viz., perturbing every pixel of every frame. In contrast, recent advances show that gradient-free selective perturbations (i.e., attacking only selected pixels and frames) could be a more realistic adversary. However, these attacks treat every frame in isolation, ignoring the relationship between neighboring states of a Markov decision process; thus resulting in high computational complexity that tends to limit their real-world plausibility due to the tight time constraint in RL. Given the above, this article showcases the first study of how transferability across frames could be exploited for boosting the creation of minimal yet powerful attacks in image-based RL. To this end, we introduce three types of frame-correlation transfers (FCTs) (i.e., anterior case transfer, random projection-based transfer, and principal components-based transfer) with varying degrees of computational complexity in generating adversaries via a genetic algorithm. We empirically demonstrate the tradeoff between the complexity and potency of the transfer mechanism by exploring four fully trained state-of-the-art policies on six Atari games. Our FCTs dramatically speed up the attack generation compared to existing methods, often reducing the computation time required to nearly zero; thus, shedding light on the real threat of real-time attacks in RL.

It is our pleasure to introduce this special issue on multitask evolutionary computation (MTEC), focusing on novel methodologies and applications of evolutionary algorithms (EAs) crafted to perform multiple search and optimization tasks jointly. EAs are population-based methods inspired by principles of natural evolution that have provided a gradient-free path to solving complex learning and optimization problems. However, unlike the natural world where evolution has engendered diverse species and produced differently skilled subpopulations,in silicoEAs are typically designed to evolve a set of solutions specialized for just a single target task. This convention of problem solving in isolation tends to curtail the power ofimplicit parallelismof a population. Skills evolved for a given problem instance do not naturally transfer to populations tasked to solve another. Hence, convergence rates remain restrained, even in settings where related tasks with overlapping search spaces, similar optimal solutions, or with other forms of reusable information, are routinely recurring.

To prevent the population from getting stuck in local areas and then missing the constrained Pareto front fragments in dealing with constrained multiobjective optimization problems (CMOPs), it is important to guide the population to evenly explore the promising areas that are not dominated by all examined feasible solutions. To this end, we first introduce a cost value-based distance into the objective space, and then use this distance and the constraints to define an indicator to evaluate the contribution of each individual to exploring the promising areas. Theoretical studies show that the proposed indicator can effectively guide population to focus on exploring the promising areas without crowding in local areas. Accordingly, we propose a new constraint handling technique (CHT) based on this indicator. To further improve the diversity of population in the promising areas, the proposed indicator-based CHT divides the promising areas into multiple subregions, and then gives priority to removing the individuals with the worst fitness values in the densest subregions. We embed the indicator-based CHT in evolutionary algorithm and propose an indicator-based constrained multiobjective algorithm for solving CMOPs. Numerical experiments on several benchmark suites show the effectiveness of the proposed algorithm. Compared with six state-of-the-art constrained evolutionary multiobjective optimization algorithms, the proposed algorithm performs better in dealing with different types of CMOPs, especially in those problems that the individuals are easy to appear in the local infeasible areas that dominate the constrained Pareto front fragments.

Many decisions in a machine learning (ML) pipeline involve nondifferentiable and discontinuous objectives and search spaces. Examples include feature selection, model selection, and hyperparameter tuning, where candidate solutions in an outer optimization loop must be evaluated via a learning subsystem. Evolutionary algorithms (EAs) are prominent gradient-free methods to handle such tasks. However, EAs are known to pose steep computational challenges, especially when dealing with large-instance datasets. As opposed to prior works that often fall back on parallel computing hardware to resolve this big data problem of EAs, in this article, we propose a novel algorithm-centric solution based onevolutionary multitasking. Our approach involves the creation of a band ofminions, i.e., small data proxies to the main target task, that are constructed by subsampling a fraction of the large dataset. We then combine the minions with the main task in a single multitask optimization framework, boosting evolutionary search by using small data to quickly optimize for the large dataset. Our key algorithmic contribution in this setting is to allocate computational resources to each of the tasks in a principled manner. The article considers wrapper-based feature selection as an illustrative case study of the broader idea of using multitasking to speedup outer loop evolutionary configurations of any ML subsystem. The experiments reveal that multitasking can indeed speedup baseline EAs, by more than 40% on some datasets.

The Vehicle Routing Problem (VRP) is a well-known NP-hard combinatorial optimization problem, which has wide spread applications in real world, such as logistics, bus route planning, and urban path planning. To solve VRP, traditional optimization methods usually start the search from scratch and ignore the VRPs solved in the past, which could lead to repeated explorations of the search space of related problems, and thus results in slow optimization process involving unnecessary computational cost. Keeping this in mind, to speed up the optimization for vehicle routing, this article presents a new study towards faster vehicle routing by transferring knowledge from customer representations which are learned from past solved VRPs. In particular, we propose to capture the useful traits buried in previous optimized routing solutions by learning a new customer representation, which can be transferred across VRPs, serving as the prior knowledge, to bias the optimization in the target VRP. In contrast to existing approaches, the proposed knowledge transfer is consist of a learning of new customer representation based on the optimized routing solution, which is general to VRPs possessing different structural properties, and a weighted$l_{1}$norm-regularized formulation for building sparse mapping across VRPs, that is easy to solve. Further, the proposed knowledge transfer across VRPs occurs along the whole optimization search process, and is thus able to guide the routing optimization process consistently. To verify the efficacy of the proposed method, by using population-based optimization method as the VRP solver, comprehensive empirical studies on both commonly used VRP benchmarks and real world vehicle routing application are presented.

Transfer covariance function, which can model domain similarity and adaptively control the knowledge transfer across domains, is widely used in transfer learning. In this paper, we concentrate on Gaussian process (GP) models using a transfer covariance function for regression problems in a black-box learning scenario. Precisely, we investigate a family of rather general transfer covariance functions,${T}_{*}$, that can model the heterogeneous sub-similarities of domains through multiple kernel learning. A necessary and sufficient condition to obtain validGPs using${T}_{*}$($GP_{T_{*}}$) for any data is given. This condition becomes specially handy for practical applications as (i) it enables semantic interpretations of the sub-similarities and (ii) it can readily be used for model learning. In particular, we propose a computationally inexpensive model learning rule that can explicitly capture different sub-similarities of domains. We propose two instantiations of$GP_{T_{*}}$, one with a set of predefined constant base kernels and one with a set of learnable parametric base kernels. Extensive experiments on 36 synthetic transfer tasks and 12 real-world transfer tasks demonstrate the effectiveness of$GP_{T_{*}}$on the sub-similarity capture and the transfer performance.

The ability to read, reason, and infer lies at the heart of neural reasoning architectures. After all, the ability to perform logical reasoning over language remains a coveted goal of Artificial Intelligence. To this end, models such as the Turing-complete differentiable neural computer (DNC) boast of real logical reasoning capabilities, along with the ability to reason beyond simple surface-level matching. In this brief, we propose the first probe into DNC's logical reasoning capabilities with a focus on text-based question answering (QA). More concretely, we propose a conceptually simple but effective adversarial attack based on metamorphic relations. Our proposed adversarial attack reduces DNCs' state-of-the-art accuracy from 100% to 1.5% in the worst case, exposing weaknesses and susceptibilities in modern neural reasoning architectures. We further empirically explore possibilities to defend against such attacks and demonstrate the utility of our adversarial framework as a simple scalable method to improve model adversarial robustness.

High performance but computationally expensive Convolutional Neural Networks (CNNs) require both algorithmic and custom hardware improvement to reduce model size and to improve energy efficiency for edge computing applications. Recent CNN architectures employ depthwise separable convolution to reduce the total number of weights and MAC operations. However, depthwise separable convolution workload does not run efficiently in existing CNN accelerators. This paper proposes an energy-efficient CONV unit for pointwise and depthwise operation. The CONV unit utilizes weight stationary to enable high efficiency. The row partial sum reduction is engaged to increase parallelism in pointwise convolution thereby lightening the memory requirements on output partial sums. Our design achieves a maximum efficiency of 3.17 TOPS/W at 0.85V/40nm CMOS which is well-suited for energy constrained edge computing applications.

In this paper, we present Conjoint Attentions (CAs), a class of novel learning-to-attend strategies for graph neural networks (GNNs). Besides considering the layer-wise node features propagated within the GNN, CAs can additionally incorporate various structural interventions, such as node cluster embedding, and higher-order structural correlations that can be learned outside of GNN, when computing attention scores. The node features that are regarded as significant by the conjoint criteria are therefore more likely to be propagated in the GNN. Given the novel Conjoint Attention strategies, we then propose Graph conjoint attention networks (CATs) that can learn representations embedded with significant latent features deemed by the Conjoint Attentions. Besides, we theoretically validate the discriminative capacity of CATs. CATs utilizing the proposed Conjoint Attention strategies have been extensively tested in well-established benchmarking datasets and comprehensively compared with state-of-the-art baselines. The obtained notable performance demonstrates the effectiveness of the proposed Conjoint Attentions.

Contrastive self-supervised learning has largely narrowed the gap to supervised pre-training on ImageNet. However, its success highly relies on the object-centric priors of ImageNet, i.e., different augmented views of the same image correspond to the same object. Such a heavily curated constraint becomes immediately infeasible when pre-trained on more complex scene images with many objects. To overcome this limitation, we introduce Object-level Representation Learning (ORL), a new self-supervised learning framework towards scene images. Our key insight is to leverage image-level self-supervised pre-training as the prior to discover object-level semantic correspondence, thus realizing object-level representation learning from scene images. Extensive experiments on COCO show that ORL significantly improves the performance of self-supervised learning on scene images, even surpassing supervised ImageNet pre-training on several downstream tasks. Furthermore, ORL improves the downstream performance when more unlabeled scene images are available, demonstrating its great potential of harnessing unlabeled data in the wild. We hope our approach can motivate future research on more general-purpose unsupervised representation learning from scene data.

Activation functions such as hyperbolic tangent (tanh) and logistic sigmoid (sigmoid) are critical computing elements in a long short term memory (LSTM) cell and network. These activation functions are non-linear, leading to challenges in their hardware implementations. Area-efficient and high performance hardware implementation of these activation functions thus becomes crucial to allow high throughput in a LSTM accelerator. In this work, we propose an approximation scheme which is suitable for both tanh and sigmoid functions. The proposed hardware for sigmoid function is 8.3 times smaller than the state-of-the-art, while for tanh function, it is the second smallest design. When applying the approximated tanh and sigmoid of 2% error in a LSTM cell computation, its final hidden state and cell state record errors of 3.1% and 5.8% respectively. When the same approximated functions are applied to a single layer LSTM network of 64 hidden nodes, the accuracy drops by 2.8% only. This proposed small yet accurate activation function hardware is promising to be used in Internet of Things (IoT) applications where accuracy can be traded off for ultra-low power consumption.

Network cluster analysis is of great importance as it is closely related to diverse applications, such as social community detection, biological module identification, and document segmentation. Aiming to effectively uncover clusters in the network data, a number of computational approaches , which utilize network topology , single vector of vertex features, or both the aforementioned, have been proposed. However, most prevalent approaches are incapable of dealing with those contemporary network data whose vertices are characterized by features collected from multiple sources. To address this challenge, in this paper, we propose a novel framework, dubbed Multi-Source Propagation Aware Network Clustering (MSPANC) for uncovering clusters in network data possessing multiple sources of vertex features. Different from most previous approaches, MSPANC is able to infer the cluster preference for each vertex utilizing both network topology and multi-source vertex features. To improve the practical significance of the discovered clusters, the learning of cluster membership is also involved into the modeling of the maximization of intra-cluster propagation regarding multi-source features. We propose a unified objective function for MSPANC to perform the clustering task and derive an alternative manner of learning algorithm for model optimization. Besides, we theoretically prove the convergence of the algorithm for optimizing MSPANC. The proposed model has been tested on five real-world datasets, including social, biological and document networks, and has been compared with several competitive baselines. The remarkable experimental results validate the effectiveness of MSPANC.

Black-box optimization algorithms typically start a search from scratch, assuming little prior knowledge about the task at hand. In practice, this approach can be prohibitive for computationally expensive problems, as a large number of costly function evaluations are often needed before a suitable (near-optimal) solution is found. Under this observation, recent efforts have incorporatedtransfer learningcapabilities into sequential model-based Bayesian optimization (BO) solvers, resulting in substantial performance speed-ups by leveraging information from related past problems. However, a common simplifying assumption in existing approaches is that the search spaces of a previously encounteredsourceand the ongoingtargettask bear the same features and dimensionality, with the difference lying in their respective objective functions. In this article, we present a generalizedtransfer BOalgorithm that relaxes the aforementioned assumption. Our method jointly transforms source features while training probabilistic transfer regression models for the target, thus applying to practical use-cases where (in addition to the difference in objective functions) the number of features could change across the source and target tasks; for example, features can be added and/or removed. The theoretical basis of our proposal is analyzed, and its empirical performance is demonstrated on synthetic benchmark functions as well as in realistic examples spanning engineering design and the automated configuration of a machine learning model. Note to Practitioners—Problems of industrial interest have a tendency of being repetitive in nature. For this reason, domain experts are always in high demand, as they are able to harness their experience of similar problems to come up with fast solutions in difficult situations. However, domain experts are not easy to find. Given this fact, the present paper puts forth a method for automating the process of knowledge extraction (through experiential learning) and transfer across problems in the domain of computationally expensive black-box optimization. The key novelty and motivation of this work lies in enabling the adaptive transfer of knowledge even when the number of features changes across the source and target problems. Our proposed approach is verified experimentally on a range of benchmarks as well as real-world problems of a computationally expensive nature, highlighting the utility of an optimization engine that is able to learn from experience without the need for constant human intervention.

Humans have the ability to identify recurring patterns in diverse situations encountered over a lifetime, constantly understanding relationships between tasks and efficiently solving them through knowledge reuse. The capacity of artificial intelligence systems to mimic such cognitive behaviors for effective problem solving is deemed invaluable, particularly when tackling real-world problems where speed and accuracy are critical. Recently, the notion of evolutionary multitasking has been explored as a means of solving multiple optimization tasks simultaneously using a single population of evolving individuals. In the presence of similarities (or even partial overlaps) between high-quality solutions of related optimization problems, the resulting scope for intertask genetic transfer often leads to significant performance speedup-as the cost of re-exploring overlapping regions of the search space is reduced. While multitasking solvers have led to recent success stories, a known shortcoming of existing methods is their inability to adapt the extent of transfer in a principled manner. Thus, in the absence of any prior knowledge about the relationships between optimization functions, a threat of predominantly negative (harmful) transfer prevails. With this in mind, this article presents a realization of a cognizant evolutionary multitasking engine within the domain of multiobjective optimization. Our proposed algorithm learns intertask relationships based on overlaps in the probabilistic search distributions derived from data generated during the course of multitasking-and accordingly adapts the extent of genetic transfers online. The efficacy of the method is substantiated on multiobjective benchmark problems as well as a practical case study of knowledge transfers from low-fidelity optimization tasks to substantially reduce the cost of high-fidelity optimization.

Objectives optimization and constraints satisfaction are two equally important goals to solve constrained many-objective optimization problems (CMaOPs). However, most existing studies for CMaOPs can be classified as feasibility-driven-constrained many-objective evolutionary algorithms (CMaOEAs), and they always give priority to satisfy constraints, while ignoring the maintenance of the population diversity for dealing with conflicting objectives. Consequently, the population may be pushed toward some locally feasible optimal or locally infeasible areas in the high-dimensional objective space. To alleviate this issue, this article presents a problem transformation technique, which transforms a CMaOP into a dynamic CMaOP (DCMaOP) for handling constraints and optimizing objectives simultaneously, to help the population cross the large and discrete infeasible regions. The well-known reference-point-based NSGA-III is tailored under the problem transformation model to solve CMaOPs, namely, DCNSGA-III. In this article, ε -feasible solutions play an important role in the proposed algorithm. To this end, in DCNSGA-III, a mating selection mechanism and an environmental selection operator are designed to generate and choose high-quality ε -feasible offspring solutions, respectively. The proposed algorithm is evaluated on a series of benchmark CMaOPs with three, five, eight, ten, and 15 objectives and compared against six state-of-the-art CMaOEAs. The experimental results indicate that the proposed algorithm is highly competitive for solving CMaOPs.

Gradient-based method has been extensively used in today's multiagent reinforcement learning (MARL). In a gradient-based MARL algorithm, each agent updates its parameterized strategy in the direction of the gradient of some performance index. However, studies on the convergence of the existing gradient-based MARL algorithms for identical interest games are quite few. In this article, we propose a policy gradient potential (PGP) algorithm that takes PGP as the source of information for guiding the strategy update, as opposed to the gradient itself, to learn the optimal joint strategy that has a maximal global reward. Since the payoff matrix and the joint strategy are often unavailable to the learning agents in reality, we consider the probability of obtaining the maximal reward as the performance index. Theoretical analysis of the PGP algorithm on the continuous model involving an identical interest repeated game shows that if the component action of every optimal joint action is unique, the critical points corresponding to all optimal joint actions are asymptotically stable. The PGP algorithm is experimentally studied and compared against other MARL algorithms on two commonly used collaborative tasks-the robots leaving a room task and the distributed sensor network task, as well as a real-world minefield navigation problem where only local state and local reward information are available. The results show that the PGP algorithm outperforms the other algorithms in terms of the cumulative reward and the number of time steps used in an episode.

The design of the evolutionary algorithm with learning capability from past search experiences has attracted growing research interests in recent years. It has been demonstrated that the knowledge embedded in the past search experience can greatly speed up the evolutionary process if properly harnessed. Autoencoding evolutionary search (AEES) is a recently proposed search paradigm, which employs a single-layer denoising autoencoder to build the mapping between two problems by configuring the solutions of each problem as the input and output for the autoencoder, respectively. The learned mapping makes it possible to perform knowledge transfer across heterogeneous problem domains with diverse properties. It has shown a promising performance of learning and transferring the knowledge from past search experiences to facilitate the evolutionary search on a variety of optimization problems. However, despite the success enjoyed by AEES, the linear autoencoding model cannot capture the nonlinear relationship between the solution sets used in the mapping construction. Taking this cue, in this article, we devise a kernelized autoencoder to construct the mapping in a reproducing kernel Hilbert space (RKHS), where the nonlinearity among problem solutions can be captured easily. Importantly, the proposed kernelized autoencoding method also holds a closed-form solution which will not bring much computational burden in the evolutionary search. Furthermore, a kernelized autoencoding evolutionary-search (KAES) paradigm is proposed that adaptively selects the linear and kernelized autoencoding along the search process in pursuit of effective knowledge transfer across problem domains. To validate the efficacy of the proposed KAES, comprehensive empirical studies on both benchmark multiobjective optimization problems as well as real-world vehicle crashworthiness design problem are presented.

Heteroscedastic regression considering the varying noises among observations has many applications in the fields, such as machine learning and statistics. Here, we focus on the heteroscedastic Gaussian process (HGP) regression that integrates the latent function and the noise function in a unified nonparametric Bayesian framework. Though showing remarkable performance, HGP suffers from the cubic time complexity, which strictly limits its application to big data. To improve the scalability, we first develop a variational sparse inference algorithm, named VSHGP, to handle large-scale data sets. Furthermore, two variants are developed to improve the scalability and capability of VSHGP. The first is stochastic VSHGP (SVSHGP) that derives a factorized evidence lower bound, thus enhancing efficient stochastic variational inference. The second is distributed VSHGP (DVSHGP) that follows the Bayesian committee machine formalism to distribute computations over multiple local VSHGP experts with many inducing points and adopts hybrid parameters for experts to guard against overfitting and capture local variety. The superiority of DVSHGP and SVSHGP compared to the existing scalable HGP/homoscedastic GP is then extensively verified on various data sets.

A key challenge in many applications of multisource transfer learning is to explicitly capture the diverse source-target similarities. In this article, we are concerned with stretching the set of practical approaches based on Gaussian process (GP) models to solve multisource transfer regression problems. Precisely, we first investigate the feasibility and performance of a family of transfer covariance functions that represent the pairwise similarity of each source and the target domain. We theoretically show that using such a transfer covariance function for general GP modeling can only capture the same similarity coefficient for all the sources, and thus may result in unsatisfactory transfer performance. This outcome, together with the scalability issues of a single GP based approach, leads us to propose TCMSStack, an integrated framework incorporating a separate transfer covariance function for each source and stacking. Contrary to typical stacking approaches, TCMSStack learns the source-target similarity in each base GP model by considering the dependencies of the other sources along the process. We introduce two instances of the proposed TCMSStack. Extensive experiments on one synthetic and two real-world data sets, with learning settings up to 11 sources for the latter, demonstrate the effectiveness of our approach.

This article embarks a study on multiagent transfer learning (TL) for addressing the specific challenges that arise in complex multiagent systems where agents have different or even competing objectives. Specifically, beyond the essential backbone of a state-of-the-art evolutionary TL framework (eTL), this article presents the novel TL framework with prediction (eTL-P) as an upgrade over existing eTL to endow agents with abilities to interact with their opponents effectively by building candidate models and accordingly predicting their behavioral strategies. To reduce the complexity of candidate models, eTL-P constructs a monotone submodular function, which facilitates to select Top-${K}$models from all available candidate models based on their representativeness in terms of behavioral coverage as well as reward diversity. eTL-P also integrates social selection mechanisms for agents to identify their better-performing partners, thus improving their learning performance and reducing the complexity of behavior prediction by reusing useful knowledge with respect to their partners’ mind universes. Experiments based on a partner-opponent minefield navigation task (PO-MNT) have shown that eTL-P exhibits the superiority in achieving higher learning capability and efficiency of multiple agents when compared to the state-of-the-art multiagent TL approaches.

Stream classification models for non-stationary environments often assume the immediate availability of data labels. However, in a practical scenario, it is quite natural that the data labels are available only after some temporal lag. This paper explores how a stream classifier model can be made adaptive to such label latency scenario. We propose SkipE-RNN, a self-evolutionary recurrent neural network with dynamically evolving skipped-recurrent-connection for the best utilization of previously observed label information while classifying the current data. When the data label is unavailable, SkipE-RNN uses an auto-learned mapping function to find the best match from the already known data labels and updates the network parameter accordingly. Later, upon availability of true data label, if the previously mapped label is found to be incorrect, SkipE-RNN employs a regularization technique along with the parameter updating process, so as to penalize the model. In addition, SkipE-RNN has inborn power of self-adjusting the network capacity by growing/pruning hidden nodes to cope with the evolving nature of data stream. Rigorous empirical evaluations using synthetic as well as real-world datasets reveal effectiveness of SkipE-RNN in both finitely delayed and infinitely delayed data label scenarios.

Adversarial perturbations are imperceptible changes to input pixels that can change the prediction of deep learning models. Learned weights of models robust to such perturbations are previously found to be transferable across different tasks but this applies only if the model architecture for the source and target tasks is the same. Input gradients characterize how small changes at each input pixel affect the model output. Using only natural images, we show here that training a student model's input gradients to match those of a robust teacher model can gain robustness close to a strong baseline that is robustly trained from scratch. Through experiments in MNIST, CIFAR-10, CIFAR-100 and Tiny-ImageNet, we show that our proposed method, input gradient adversarial matching, can transfer robustness across different tasks and even across different model architectures. This demonstrates that directly targeting the semantics of input gradients is a feasible way towards adversarial robustness.

Joint clustering and feature learning methods have shown remarkable performance in unsupervised representation learning. However, the training schedule alternating between feature clustering and network parameters update leads to unstable learning of visual representations. To overcome this challenge, we propose Online Deep Clustering (ODC) that performs clustering and network update simultaneously rather than alternatingly. Our key insight is that the cluster centroids should evolve steadily in keeping the classifier stably updated. Specifically, we design and maintain two dynamic memory modules, i.e., samples memory to store samples' labels and features, and centroids memory for centroids evolution. We break down the abrupt global clustering into steady memory update and batch-wise label re-assignment. The process is integrated into network update iterations. In this way, labels and the network evolve shoulder-to-shoulder rather than alternatingly. Extensive experiments demonstrate that ODC stabilizes the training process and boosts the performance effectively.

Object detection, which aims at recognizing or locating the objects of interest in remote sensing imagery with high spatial resolutions (HSR), plays a significant role in many real-world scenarios, e.g., environment monitoring, urban planning, civil infrastructure construction, disaster rescuing, and geographic image retrieval . As a long-lasting challenging problem in both machine learning and geoinformatics communities, many approaches have been proposed to tackle it. However, previous methods always overlook the abundant information embedded in the HSR remote sensing images. The effectiveness of these methods, e.g., accuracy of detection, is therefore limited to some extent. To overcome the mentioned challenge, in this paper, we propose a novel two-phase deep framework, dubbed GLGOD-Net, to effectively detect meaningful objects in HSR images . GLGOD-Net firstly attempts to learn the enhanced deep representations from super-resolution image data . Fully utilizing the augmented image representations, GLGOD-Net then learns the fused representations into which both local and global latent features are implanted. Such fused representations learned by GLGOD-Net can be used to precisely detect different objects in remote sensing images . The proposed framework has been extensively tested on a real-world HSR image dataset for object detection and has been compared with several strong baselines. The remarkable experimental results validate the effectiveness of GLGOD-Net. The success of GLGOD-Net not only advances the cutting-edge of image data analytics , but also promotes the corresponding applicability of deep learning in remote sensing imagery .

Graph clustering, which aims at discovering sets of related vertices in graph-structured data, plays a crucial role in various applications, such as social community detection and biological module discovery. With the huge increase in the volume of data in recent years, graph clustering is used in an increasing number of real-life scenarios. However, the classical and state-of-the-art methods, which consider only single-view features or a single vector concatenating features from different views and neglect the contextual correlation between pairwise features, are insufficient for the task, as features that characterize vertices in a graph are usually from multiple views and the contextual correlation between pairwise features may influence the cluster preference for vertices. To address this challenging problem, we introduce in this paper, a novel graph clustering model, dubbed contextual correlation preserving multiview featured graph clustering (CCPMVFGC) for discovering clusters in graphs with multiview vertex features. Unlike most of the aforementioned approaches, CCPMVFGC is capable of learning a shared latent space from multiview features as the cluster preference for each vertex and making use of this latent space to model the inter-relationship between pairwise vertices. CCPMVFGC uses an effective method to compute the degree of contextual correlation between pairwise vertex features and utilizes view-wise latent space representing the feature-cluster preference to model the computed correlation. Thus, the cluster preference learned by CCPMVFGC is jointly inferred by multiview features, view-wise correlations of pairwise features, and the graph topology. Accordingly, we propose a unified objective function for CCPMVFGC and develop an iterative strategy to solve the formulated optimization problem. We also provide the theoretical analysis of the proposed model, including convergence proof and computational complexity analysis. In our experiments, we extensively compare the proposed CCPMVFGC with both classical and state-of-the-art graph clustering methods on eight standard graph datasets (six multiview and two single-view datasets). The results show that CCPMVFGC achieves competitive performance on all eight datasets, which validates the effectiveness of the proposed model.

Humans rarely tackle every problem from scratch. Given this observation, the motivation for this paper is to improve optimization performance through adaptive knowledge transfer across related problems. The scope for spontaneous transfers under the simultaneous occurrence of multiple problems unveils the benefits of multitasking. Multitask optimization has recently demonstrated competence in solving multiple (related) optimization tasks concurrently. Notably, in the presence of underlying relationships between problems, the transfer of high-quality solutions across them has shown to facilitate superior performance characteristics. However, in the absence of any prior knowledge about the intertask synergies (as is often the case with general black-box optimization), the threat of predominantly negative transfer prevails. Susceptibility to negative intertask interactions can impede the overall convergence behavior. To allay such fears, in this paper, we propose a novel evolutionary computation framework that enables online learning and exploitation of the similarities (and discrepancies) between distinct tasks in multitask settings, for an enhanced optimization process. Our proposal is based on the principled theoretical arguments that seek to minimize the tendency of harmful interactions between tasks, based on a purely data-driven learning of relationships among them. The efficacy of our proposed method is validated experimentally on a series of synthetic benchmarks, as well as a practical study that provides insights into the behavior of the method in the face of several tasks occurring at once.

Structured multioutput learning is a topic in artificial intelligence that considers multiple structured outputs prediction for a given input. The output may involve structured objects in the form of sequence, string, tree, lattice, or graph and has values that are characterized by diverse data types, such as binary, nominal, ordinal, and real-valued variables. Such learning problems arise in a variety of real-world applications, ranging from document classification, computer emulation, sensor network analysis, concept-based information retrieval, and human action/causal induction to video analysis, image annotation/retrieval, gene function prediction, and brain science. As many complex real-world scenarios can be posed as a structured multioutput learning problem, their importance and popularity have been increasing steadily.

The vast quantity of information brought by big data as well as the evolving computer hardware encourages success stories in the machine learning community. In the meanwhile, it poses challenges for the Gaussian process regression (GPR), a well-known nonparametric, and interpretable Bayesian model, which suffers from cubic complexity to data size. To improve the scalability while retaining desirable prediction quality, a variety of scalable GPs have been presented. However, they have not yet been comprehensively reviewed and analyzed to be well understood by both academia and industry. The review of scalable GPs in the GP community is timely and important due to the explosion of data size. To this end, this article is devoted to reviewing state-of-the-art scalable GPs involving two main categories: global approximations that distillate the entire data and local approximations that divide the data for subspace learning. Particularly, for global approximations, we mainly focus on sparse approximations comprising prior approximations that modify the prior but perform exact inference, posterior approximations that retain exact prior but perform approximate inference, and structured sparse approximations that exploit specific structures in kernel matrix; for local approximations, we highlight the mixture/product of experts that conducts model averaging from multiple local experts to boost predictions. To present a complete review, recent advances for improving the scalability and capability of scalable GPs are reviewed. Finally, the extensions and open issues of scalable GPs in various scenarios are reviewed and discussed to inspire novel ideas for future research avenues.

The aim of multi-output learning is to simultaneously predict multiple outputs given an input. It is an important learning problem for decision-making since making decisions in the real world often involves multiple complex factors and criteria. In recent times, an increasing number of research studies have focused on ways to predict multiple outputs at once. Such efforts have transpired in different forms according to the particular multi-output learning problem under study. Classic cases of multi-output learning include multi-label learning, multi-dimensional learning, multi-target regression, and others. From our survey of the topic, we were struck by a lack in studies that generalize the different forms of multi-output learning into a common framework. This article fills that gap with a comprehensive review and analysis of the multi-output learning paradigm. In particular, we characterize the four Vs of multi-output learning, i.e., volume, velocity, variety, and veracity, and the ways in which the four Vs both benefit and bring challenges to multi-output learning by taking inspiration from big data. We analyze the life cycle of output labeling, present the main mathematical definitions of multi-output learning, and examine the field's key challenges and corresponding solutions as found in the literature. Several model evaluation metrics and popular data repositories are also discussed. Last but not least, we highlight some emerging challenges with multi-output learning from the perspective of the four Vs as potential research directions worthy of further studies.

Genetic programming (GP) is a powerful evolutionary algorithm that has been widely used for solving many real-world optimization problems. However, traditional GP can only solve a single task in one independent run, which is inefficient in cases where multiple tasks need to be solved at the same time. Recently, multifactorial optimization (MFO) has been proposed as a new evolutionary paradigm toward evolutionary multitasking. It intends to conduct evolutionary search on multiple tasks in one independent run. To enable multitasking GP, in this paper, we propose a novel multifactorial GP (MFGP) algorithm. To the best of our knowledge, this is the first attempt in the literature to conduct multitasking GP using a single population. The proposed MFGP consists of a novel scalable chromosome encoding scheme which is capable of representing multiple solutions simultaneously, and new evolutionary mechanisms for MFO based on self-learning gene expression programming. Further, comprehensive experimental studies are conducted on multitask scenarios consisting of commonly used GP benchmark problems and real world applications. The obtained empirical results confirmed the efficacy of the proposed MFGP.

Sample-based evolutionary algorithms (EAs) are widely used for optimizing problems with multi (greater than one but less than four) or even many (greater than or equal to four) objectives of interest. In general, the difficulty of a problem exponentially increases with the number of objectives, serving as a clear example of the curse of dimensionality. The exploratory approach an EA takes in these cases has led to it being thought of as a big data generator, progressively sampling and evaluating solutions in high performing regions of a decision space to guide the search towards optimal solutions. Notably, in both multi- and many-objective EAs, the sampled data can be further utilized for building inverse generative models, mapping points in objective space back to solutions in the decision space. Such models offer immense flexibility to a decision maker in generating new target solutions on the fly, thereby facilitating real-time a posteriori preference incorporation into the search. In this paper, we show that the data distribution resulting from a many-objective formulation is in fact more conducive to building accurate inverse models than its multiobjective counterpart. Given the potential utility of these models, we in turn shed light on a rare blessing of dimensionality that is yet to be explored in the context of optimization. We first present simple theoretical arguments supporting our claim. Thereafter, experimental studies of Gaussian process-based inverse modeling for a synthetic and a real-world example are carried out to further confirm the theory.

Transfer optimization is an emerging concept that promises to enhance productivity of planning and decision-making processes by allowing for the adaptive reuse of knowledge (data) drawn from various “source” problems in a related ongoing “target” task of interest. Despite the recent advances in transfer optimization, however, a continuing challenge is the scalability of associated algorithms given big data of source problem instances. This paper tackles the scaling problem of an online adaptive knowledge transfer framework under big source data. We propose an efficient source selection algorithm based on the theory of multi-armed bandits such that the most related source task to the target is chosen for knowledge transfer, as opposed to extracting knowledge from all sources simultaneously. For this purpose, we introduce a novel and principled reward measure to reflect the source-target similarities. The efficacy of our proposed approach is assessed on the well-known knapsack problem that has practical implications in optimization of supply chain and manufacturing processes. Extensive experiments are conducted under big data of source problem instances. The numerical results clearly reveal that the incorporation of the proposed source selection mechanism in the existing adaptive knowledge transfer framework makes it successfully feasible for fast/real-time decision-making in the big data source setting.

Neuroevolution (i.e., training neural network with Evolution Computation) has successfully unfolded a range of challenging reinforcement learning (RL) tasks. However, existing neuroevolution methods suffer from high sample complexity, as the black-box evaluations (i.e., accumulated rewards of complete Markov Decision Processes (MDPs)) discard bunches of temporal frames (i.e., time-step data instances in MDP). Actually, these temporal frames hold the Markov property of the problem, that benefits the training of neural network as well by temporal difference (TD) learning. In this paper, we propose a memetic reinforcement learning (MRL) framework that optimizes the RL agent by leveraging both black-box evaluations and temporal frames. To this end, an evolution strategy (ES) is associated with Q learning, where ES provides diversified frames to globally train the agent, and Q learning locally exploits the Markov property within frames to refresh the agent. Therefore, MRL conveys a novel memetic framework that allows evaluation free local search by Q learning. Experiments on classical control problem verify the efficiency of the proposed MRL, that achieves significantly faster convergence than canonical ES.

Autonomous construction of deep neural network (DNNs) is desired for data streams because it potentially offers two advantages: proper model's capacity and quick reaction to drift and shift. While self-organizing mechanism of DNNs remains an open issue, this task is even more challenging to be developed for standard multi-layer DNNs than that using the different-depth structures, because addition of a new layer results in information loss of previously trained knowledge. A Neural Network with Dynamically Evolved Capacity (NADINE) is proposed in this paper. NADINE features a fully open structure where its network structure, depth and width, can be automatically evolved from scratch in the online manner and without the use of problem-specific thresholds. NADINE is structured under a standard MLP architecture and the catastrophic forgetting issue during the hidden layer addition phase is resolved using the proposal of soft-forgetting and adaptive memory methods. The advantage of NADINE, namely elastic structure and online learning trait, is numerically validated using nine data stream classification and regression problems where it demonstrates performance's improvement over prominent algorithms in all problems. In addition, it is capable of dealing with data stream regression and classification problems equally well.

Stochastic block models (SBMs) play essential roles in network analysis, especially in those related to unsupervised learning (clustering). Many SBM-based approaches have been proposed to uncover network clusters, by means of maximizing the block-wise posterior probability that generates edges bridging vertices. However, none of them is capable of inferring the cluster preference for each vertex through simultaneously modeling block-wise edge structure, vertex features, and similarities between pairwise vertices. To fill this void, we propose a novel SBM dubbed manifold regularized stochastic model (MrSBM) to perform the task of unsupervised learning in network data in this paper. Besides modeling edges that are within or connecting blocks, MrSBM also considers modeling vertex features utilizing the probabilities of vertex-cluster preference and feature-cluster contribution. In addition, MrSBM attempts to generate manifold similarity of pairwise vertices utilizing the inferred vertex-cluster preference. As a result, the inference of cluster preference may well capture the comparability in the manifold. We design a novel process for network data generation, based on which, we specify the model structure and formulate the network clustering problem using a novel likelihood function. To guarantee MrSBM learns the optimal cluster preference for each vertex, we derive an effective Expectation-Maximization based algorithm for model fitting. MrSBM has been tested on five sets of real-world network data and has been compared with both classical and state-of-the-art approaches to network clustering. The competitive experimental results validate the effectiveness of MrSBM.

In transfer learning , we aim to improve the predictive modeling of a target output by using the knowledge from some related source outputs. In real-world applications, the data from the target domain is often precious and hard to obtain, while the data from source domains is plentiful. Thus, since the complexity of Gaussian process based multi-task/transfer learning approaches grows cubically with the total number of source＋ target observations, the method becomes increasingly impractical for large ( > 1 0 4 ) source data inputs even with a small amount of target data. In order to scale known transfer Gaussian processes to large-scale source datasets , we propose an efficient aggregation model in this paper, which combines the predictions from distributed (small-scale) local experts in a principled manner. The proposed model inherits the advantages of single-task aggregation schemes, including efficient computation, analytically tractable inference, and straightforward parallelization during training and prediction. Further, a salient feature of the proposed method is the enhanced expressiveness in transfer learning — as a byproduct of flexible inter-task relationship modelings across different experts. When deploying such models in real-world applications, each local expert corresponds to a lightweight predictor that can be embedded in edge devices, thus catering to cases of online on-mote processing in fog computing settings.

As a non-parametric Bayesian model which produces informative predictive distribution , Gaussian process (GP) has been widely used in various fields, like regression, classification and optimization. The cubic complexity of standard GP however leads to poor scalability, which poses challenges in the era of big data . Hence, various scalable GPs have been developed in the literature in order to improve the scalability while retaining desirable prediction accuracy. This paper devotes to investigating the methodological characteristics and performance of representative global and local scalable GPs including sparse approximations and local aggregations from four main perspectives: scalability, capability, controllability and robustness. The numerical experiments on two toy examples and five real-world datasets with up to 250K points offer the following findings. In terms of scalability, most of the scalable GPs own a time complexity that is linear to the training size. In terms of capability, the sparse approximations capture the long-term spatial correlations , the local aggregations capture the local patterns but suffer from over-fitting in some scenarios. In terms of controllability, we could improve the performance of sparse approximations by simply increasing the inducing size. But this is not the case for local aggregations. In terms of robustness, local aggregations are robust to various initializations of hyperparameters due to the local attention mechanism . Finally, we highlight that the proper hybrid of global and local scalable GPs may be a promising way to improve both the model capability and scalability for big data .

This paper draws motivation from the remarkable ability of humans to extract useful building-blocks of knowledge from past experiences and spontaneously reuse them for new and more challenging tasks. It is contended that successfully replicating such capabilities in computational solvers, particularly global black-box optimizers, can lead to significant performance enhancements over the current state-of-the-art. The main challenge to overcome is that in general black-box settings, no problem-specific data may be available prior to the onset of the search, thereby limiting the possibility of offline measurement of the synergy between problems. In light of the above, this paper introduces a novel evolutionary computation framework that enables online learning and exploitation of similarities across optimization problems, with the goal of achieving an algorithmic realization of the transfer optimization paradigm. One of the salient features of our proposal is that it accounts for latent similarities which while being less apparent on the surface, may be gradually revealed during the course of the evolutionary search. A theoretical analysis of our proposed framework is carried out, substantiating its positive influences on optimization performance. Furthermore, the practical efficacy of an instantiation of an adaptive transfer evolutionary algorithm is demonstrated on a series of numerical examples, spanning discrete, continuous, as well as single- and multi-objective optimization.

Evolutionary multitasking (EMT) is an emerging research topic in the field of evolutionary computation. In contrast to the traditional single-task evolutionary search, EMT conducts evolutionary search on multiple tasks simultaneously. It aims to improve convergence characteristics across multiple optimization problems at once by seamlessly transferring knowledge among them. Due to the efficacy of EMT, it has attracted lots of research attentions and several EMT algorithms have been proposed in the literature. However, existing EMT algorithms are usually based on a common mode of knowledge transfer in the form of implicit genetic transfer through chromosomal crossover. This mode cannot make use of multiple biases embedded in different evolutionary search operators, which could give better search performance when properly harnessed. Keeping this in mind, this paper proposes an EMT algorithm with explicit genetic transfer across tasks, namely EMT via autoencoding, which allows the incorporation of multiple search mechanisms with different biases in the EMT paradigm. To confirm the efficacy of the proposed EMT algorithm with explicit autoencoding, comprehensive empirical studies have been conducted on both the single- and multi-objective multitask optimization problems.

The multilayer perceptrons (MLPs) are widely used in many fields, however, singularities in the parameter space may seriously influence the learning dynamics of MLPs and cause strange learning behaviors. Given that the singularities are the subspaces of the parameter space where the Fisher information matrix (FIM) degenerates, the FIM plays a key role in the study of the singular learning dynamics of the MLPs. In this paper, we obtain the analytical form of the FIM for unipolar activation function-based MLPs where the input subjects to the Gaussian distribution with general covariance matrix and the unipolar error function is chosen as the activation function. Then three simulation experiments are taken to verify the validity of the obtained results.

This paper proposes a Gaussian process (GP) based co-sub-Pareto front surrogate augmentation strategy for evolutionary optimization of computationally expensive multiobjective problems. In the proposed algorithm, a multiobjective problem is decomposed into a number of subproblems, the solution of each of which is used to approximate a portion or sector of the Pareto front (i.e., a subPF). Thereafter, a multitask GP model is incorporated to exploit the correlations across the subproblems via joint surrogate model learning. A novel criterion for the utility function is defined on the surrogate landscape to determine the next candidate solution for evaluation using the actual expensive objectives. In addition, a new management strategy for the evaluated solutions is presented for model building. The novel feature of our approach is that it infers multiple subproblems jointly by exploiting the possible dependencies between them, such that knowledge can be transferred across subPFs approximated by the subproblems. Experimental studies under several scenarios indicate that the proposed algorithm outperforms state-of-the-art multiobjective evolutionary algorithms for expensive problems. The parameter sensitivity and effectiveness of the proposed algorithm are analyzed in detail.

Real-world applications typically have multiple sparse reconstruction tasks to be optimized. In order to exploit the similar sparsity pattern between different tasks, this paper establishes an evolutionary multitasking framework to simultaneously optimize multiple sparse reconstruction tasks using a single population. In the proposed method, the evolutionary algorithm aims to search the locations of nonzero components or rows instead of searching sparse vector or matrix directly. Then the within-task and between-task genetic transfer operators are employed to reinforce the exchange of genetic material belonging to the same or different tasks. The proposed method can solve multiple measurement vector problems efficiently because the length of decision vector is independent of the number of measurement vectors. Finally, a case study on hyperspectral image unmixing is investigated in an evolutionary multitasking setting. It is natural to consider a sparse unmixing problem in a homogeneous region as a task. Experiments on signal reconstruction and hyperspectral image unmixing demonstrate the effectiveness of the proposed multitasking framework for sparse reconstruction.

In most real-world settings, designs are often gradually adapted and improved over time. Consequently, there exists knowledge from distinct (but possibly related) design exercises, which have either been previously completed or are currently in-progress, that may be leveraged to enhance the optimization performance of a particular target optimization task of interest. Further, it is observed that modern day design cycles are typically distributed in nature, and consist of multiple teams working on associated ideas in tandem. In such environments, vast amounts of related information can become available at various stages of the search process corresponding to some ongoing target optimization exercise. Successfully exploiting this knowledge is expected to be of significant value in many practical settings, where solving an optimization problem from scratch may be exorbitantly costly or time consuming. Accordingly, in this paper, we propose an adaptive knowledge reuse framework for surrogate-assisted multiobjective optimization of computationally expensive problems, based on the novel idea of multiproblem surrogates. This idea provides the capability to acquire and spontaneously transfer learned models across problems, facilitating efficient global optimization. The efficacy of our proposition is demonstrated on a series of synthetic benchmark functions, as well as two practical case studies.

The DTLZ1-DTLZ4 problems are by far one of the most commonly used test problems in the validation and comparison of multiobjective optimization evolutionary algorithms (MOEAs). However, very recently, it has been pointed out that they have the following two special characteristics: 1) the regularly oriented Pareto front shape and 2) the single distance function. As a modification of them, this paper presents a new set of test problems mDTLZ1-mDTLZ4 to avoid the two special characteristics. Using these new test problems, we investigate the performance of three representative multiobjective evolutionary algorithms NSGA-II, MOEA/D-Tch, and SMS-EMOA. Experimental results indicate that the performance of NSGA-II and MOEA/D-Tch deteriorates on mDTLZ1-mDTLZ4. Subsequently, our analysis reveals that there exist the hardly dominated boundaries in each of mDTLZ1-mDTLZ4, which hinder the approximation of Pareto dominance-based algorithms and Tchebycheff-decomposition-based algorithms. Furthermore, we summarize that the hardly dominated boundary should be an often encountered problem feature in multiobjective optimization. Last but not least, we point out and validate some coping strategies for dominance-based algorithms and decomposition-based algorithms to overcome the challenges caused by the hardly dominated boundary.

In some real-world applications, it has been found that the performance of multiobjective optimization evolutionary algorithms (MOEAs) may deteriorate when boundary solutions in the Pareto front (PF) are more difficult to approximate than others. Such a problem feature, referred to as difficult-to-approximate (DtA) PF boundaries, is seldom considered in existing multiobjective optimization test problems. To fill this gap and facilitate possible systematic studies, we introduce a new test problem generator. The proposed generator enables the design of test problems with controllable difficulties regarding the feature of DtA PF boundaries. Three representative MOEAs, NSGA-II, SMS-EMOA, and MOEA/D-DRA, are performed on a series of test problems created using the proposed generator. Experimental results indicate that all the three algorithms perform poorly on the new test problems. Meanwhile, a modified variant of MOEA/D-DRA, denoted as MOEA/D-DRA-UT, is validated to be more effective in dealing with these problems. Subsequently, it is concluded that the rational allocation of computational resources between different PF parts is crucial for MOEAs to handle the problems with DtA PF boundaries.

In most domain adaption approaches, all features are used for domain adaption. However, often, not every feature is beneficial for domain adaption. In such cases, incorrectly involving all features might cause the performance to degrade. In other words, to make the model trained on the source domain work well on the target domain, it is desirable to find invariant features for domain adaption rather than using all features. However, invariant features across domains may lie in a higher order space, instead of in the original feature space. Moreover, the discriminative ability of some invariant features such as shared background information is weak, and needs to be further filtered. Therefore, in this paper, we propose a novel domain adaption algorithm based on an explicit feature map and feature selection. The data are first represented by a kernel-induced explicit feature map, such that high-order invariant features can be revealed. Then, by minimizing the marginal distribution difference, conditional distribution difference, and the model error, the invariant discriminative features are effectively selected. This problem is NP-hard to be solved, and we propose to relax it and solve it by a cutting plane algorithm. Experimental results on six real-world benchmarks have demonstrated the effectiveness and efficiency of the proposed algorithm, which outperforms many state-of-the-art domain adaption approaches.

Embedding methods have shown promising performance in multi-label prediction, as they can discover the dependency of labels. Most embedding methods cannot well align the input and output, which leads to degradation in prediction performance. Besides, they suffer from expensive prediction computational costs when applied to large-scale datasets. To address the above issues, this paper proposes a Co-Hashing (CoH) method by formulating multi-label learning from the perspective of cross-view learning. CoH first regards the input and output as two views, and then aims to learn a common latent hamming space, where input and output pairs are compressed into compact binary embeddings. CoH enjoys two key benefits: 1) the input and output can be well aligned, and their correlations are explored; 2) the prediction is very efficient using fast cross-view kNN search in the hamming space. Moreover, we provide the generalization error bound for our method. Extensive experiments on eight real-world datasets demonstrate the superiority of the proposed CoH over the state-of-the-art methods in terms of both prediction accuracy and efficiency.

Dynamic multi-objective optimization problem (D-MOP) is widely existed in many real-world applications. Over the years, DMOP has attracted many research attentions in the literature. The adaptive indicator-based evolutionary algorithm (IBEA2) is a recently proposed multi-objective evolutionary algorithm (MOEA). It has demonstrated strong search capability on commonly used multi-objective benchmarks over state-of-the-art MOEAs. However, as the adaptation of parameter$k$is based on the selected solutions with maximum hypervolume, this mechanism will be inappropriate if the problem changes over time. The reason is that the solutions with high hypervolume at one particular time instance may not be with high hypervolume at another if the problem changed. Keeping this in mind, inspired by the recent autoencoding evolutionary search, which is able to transfer the past search experiences to improve the evolutionary search on unseen problems, in this paper, we propose to extend the IBEA2 by adapting k with transferred high hypervolume solutions obtained before the dynamic change occurs, for solving DMOP. To evaluate the proposed method, empirical comparisons on the commonly used Farina-Deb-Amato (FDA) DMOP benchmarks, against both the IBEA2 and one recently proposed dynamic MOEA, are presented.

Transfer covariance functions, which can model domain similarities and adaptively control the knowledge transfer across domains, are widely used in Gaussian process (GP) based transfer learning. We focus on regression problems in a black-box learning scenario, and study a family of rather general transfer covariance functions, T_*, that can model the similarity heterogeneity of domains through multiple kernel learning. A necessary and sufficient condition that (i) validates GPs using T_* for any data and (ii) provides semantic interpretations is given. Moreover, building on this condition, we propose a computationally inexpensive model learning rule that can explicitly capture different sub-similarities of domains. Extensive experiments on one synthetic dataset and four real-world datasets demonstrate the effectiveness of the learned GP on the sub-similarity capture and the transfer performance.

In order to scale standard Gaussian process (GP) regression to large-scale datasets, aggregation models employ factorized training process and then combine predictions from distributed experts. The state-of-the-art aggregation models, however, either provide inconsistent predictions or require time-consuming aggregation process. We first prove the inconsistency of typical aggregations using disjoint or random data partition, and then present a consistent yet efficient aggregation model for large-scale GP. The proposed model inherits the advantages of aggregations, e.g., closed-form inference and aggregation, parallelization and distributed computing. Furthermore, theoretical and empirical analyses reveal that the new aggregation model performs better due to the consistent predictions that converge to the true underlying function when the training size approaches infinity.

kNN embedding methods, such as the state-of-the-art LM-kNN, have shown impressive results in multi-label learning. Unfortunately, these approaches suffer expensive computation and memory costs in large-scale settings. To fill this gap, this paper proposes a novel deep prototype compression, i.e., DBPC for fast multi-label prediction. DBPC compresses the database into a small set of short discrete prototypes, and uses the prototypes for prediction. The benefit of DBPC comes from two aspects: 1) The number of distance comparisons are reduced in the prototype; 2) The distance computation cost is significantly decreased in the reduced space. We propose to jointly learn the deep latent subspace and discrete prototypes within one framework. The encoding and decoding neural networks are employed to make deep discrete prototypes well represent the instances and labels. Extensive experiments on several large-scale datasets demonstrate that DBPC achieves several orders of magnitude lower storage and prediction complexity than state-of-the-art multi-label methods, while achieving competitive accuracy.

Network embedding aims to seek low-dimensional vector representations for network nodes, by preserving the network structure. The network embedding is typically represented in continuous vector, which imposes formidable challenges in storage and computation costs, particularly in large-scale applications. To address the issue, this paper proposes a novel discrete network embedding (DNE) for more compact representations. In particular, DNE learns short binary codes to represent each node. The Hamming similarity between two binary embeddings is then employed to well approximate the ground-truth similarity. A novel discrete multi-class classifier is also developed to expedite classification. Moreover, we propose to jointly learn the discrete embedding and classifier within a unified framework to improve the compactness and discrimination of network embedding. Extensive experiments on node classification consistently demonstrate that DNE exhibits lower storage and computational complexity than state-of-the-art network embedding methods, while obtains competitive classification results.

The existence of singularities often affects the learning dynamics in feedforward neural networks. In this paper, based on theoretical analysis results, we numerically analyze the learning dynamics of radial basis function (RBF) networks near singularities to understand to what extent singularities influence the learning dynamics. First, we show the explicit expression of the Fisher information matrix for RBF networks. Second, we demonstrate through numerical simulations that the singularities have a significant impact on the learning dynamics of RBF networks. Our results show that overlap singularities mainly have influence on the low dimensional RBF networks and elimination singularities have a more significant impact to the learning processes than overlap singularities in both low and high dimensional RBF networks, whereas the plateau phenomena are mainly caused by the elimination singularities. The results can also be the foundation to investigate the singular learning dynamics in deep feedforward neural networks.

Many-objective optimization problems bring great difficulties to the existing multiobjective evolutionary algorithms, in terms of selection operators, computational cost, visualization of the high-dimensional tradeoff front, and so on. Objective reduction can alleviate such difficulties by removing the redundant objectives in the original objective set, which has become one of the most important techniques in many-objective optimization. In this paper, we suggest to view objective reduction as a multiobjective search problem and introduce three multiobjective formulations of the problem, where the first two formulations are both based on preservation of the dominance structure and the third one utilizes the correlation between objectives. For each multiobjective formulation, a multiobjective objective reduction algorithm is proposed by employing the nondominated sorting genetic algorithm II to generate a Pareto front of nondominated objective subsets that can offer decision support to the user. Moreover, we conduct a comprehensive analysis of two major categories of objective reduction approaches based on several theorems, with the aim of revealing their strengths and limitations. Lastly, the performance of the proposed multiobjective algorithms is studied extensively on various benchmark problems and two real-world problems. Numerical results and comparisons are then shown to highlight the effectiveness and superiority of the proposed multiobjective algorithms over existing state-of-the-art approaches in the related field.

Embedding methods have shown promising performance in multilabel prediction, as they are able to discover the label dependence. However, most methods ignore the correlations between the input and output, such that their learned embeddings are not well aligned, which leads to degradation in prediction performance. This paper presents a formulation for multilabel learning, from the perspective of cross-view learning, that explores the correlations between the input and the output. The proposed method, called Co-Embedding (CoE), jointly learns a semantic common subspace and view-specific mappings within one framework. The semantic similarity structure among the embeddings is further preserved, ensuring that close embeddings share similar labels. Additionally, CoE conducts multilabel prediction through the cross-view $k$ nearest neighborhood ( $k$ NN) search among the learned embeddings, which significantly reduces computational costs compared with conventional decoding schemes. A hashing-based model, i.e., Co-Hashing (CoH), is further proposed. CoH is based on CoE, and imposes the binary constraint on continuous latent embeddings. CoH aims to generate compact binary representations to improve the prediction efficiency by benefiting from the efficient $k$ NN search of multiple labels in the Hamming space. Extensive experiments on various real-world data sets demonstrate the superiority of the proposed methods over the state of the arts in terms of both prediction accuracy and efficiency.

Multiobjective evolutionary algorithms (MOEAs) have proven their effectiveness and efficiency in solving problems with two or three objectives. However, recent studies show that MOEAs face many difficulties when tackling problems involving a larger number of objectives as their behavior becomes similar to a random walk in the search space since most individuals are nondominated with respect to each other. Motivated by the interesting results of decomposition-based approaches and preference-based ones, we propose in this paper a new decomposition-based dominance relation to deal with many-objective optimization problems and a new diversity factor based on the penalty-based boundary intersection method. Our reference point-based dominance (RP-dominance), has the ability to create a strict partial order on the set of nondominated solutions using a set of well-distributed reference points. The RP-dominance is subsequently used to substitute the Pareto dominance in nondominated sorting genetic algorithm-II (NSGA-II). The augmented MOEA, labeled as RP-dominance-based NSGA-II, has been statistically demonstrated to provide competitive and oftentimes better results when compared against four recently proposed decomposition-based MOEAs on commonly-used benchmark problems involving up to 20 objectives. In addition, the efficacy of the algorithm on a realistic water management problem is showcased.

Recent analytical studies have revealed that in spite of promising success in problem solving, the performance of evolutionary multitasking deteriorates with decreasing similarity between constitutive tasks. The present day multifactorial evolutionary algorithm (MFEA) is susceptible to negative knowledge transfer between uncorrelated tasks. To alleviate this issue, we propose a linearized domain adaptation (LDA) strategy that transforms the search space of a simple task to the search space similar to its constitutive complex task. This high order representative space resembles high correlation with its constitutive task and provides a platform for efficient knowledge transfer via crossover. The proposed framework, LDA-MFEA is tested on several benchmark problems constituting of tasks with different degrees of similarities and intersecting global optima. Experimental results demonstrate competitive performances against MFEA and shows that our proposition dramatically improves the performance relative to optimizing each task independently.

To date, dynamic vehicle routing problem (DVRP) has attracted great research attentions due to its wide range of real world applications. In contrast to traditional static vehicle routing problem, the whole routing information in DVRP is usually unknown and obtained dynamically during the routing execution process. To solve DVRP, many heuristic and metaheuristic methods have been proposed in the literature. In this paper, we present a novel evolutionary search paradigm with learning capability for solving DVRP. In particular, we propose to capture the structured knowledge from optimized routing solution in early time slot, which can be further reused to bias the customer-vehicle assignment when dynamic occurs. By extending our previous research work, the learning of useful knowledge, and the scheduling of dynamic customer requests are detailed here. Further, to evaluate the efficacy of the proposed search paradigm, comprehensive empirical studies on 21 commonly used DVRP instances with diverse properties are also reported.

Multi-instance multi-label (MIML) learning has many interesting applications in computer visions, including multi-object recognition and automatic image tagging. In these applications, additional information such as bounding-boxes, image captions and descriptions is often available during training phrase, which is referred as privileged information (PI). However, as existing works on learning using PI only consider instance-level PI (privileged instances), they fail to make use of bag-level PI (privileged bags) available in MIML learning. Therefore, in this paper, we propose a two-stream fully convolutional network, named MIML-FCN+, unified by a novel PI loss to solve the problem of MIML learning with privileged bags. Compared to the previous works on PI, the proposed MIML-FCN+ utilizes the readily available privileged bags, instead of hard-to-obtain privileged instances, making the system more general and practical in real world applications. As the proposed PI loss is convex and SGD-compatible and the framework itself is a fully convolutional network, MIML FCN+ can be easily integrated with state-of-the-art deep learning networks. Moreover, the flexibility of convolutional layers allows us to exploit structured correlations among instances to facilitate more effective training and testing. Experimental results on three benchmark datasets demonstrate the effectiveness of the proposed MIML-FCN+, outperforming state-of-the-art methods in the application of multi-object recognition.

A key challenge in multi-source transfer learning is to capture the diverse inter-domain similarities. In this paper, we study different approaches based on Gaussian process models to solve the multi-source transfer regression problem. Precisely, we first investigate the feasibility and performance of a family of transfer covariance functions that represent the pairwise similarity of each source and the target domain. We theoretically show that using such a transfer covariance function for general Gaussian process modelling can only capture the same similarity coefficient for all the sources, and thus may result in unsatisfactory transfer performance. This leads us to propose TC$_{MS}$Stack, an integrated strategy incorporating the benefits of the transfer covariance function and stacking. Extensive experiments on one synthetic and two real-world datasets, with learning settings of up to 11 sources for the latter, demonstrate the effectiveness of our proposed TC$_{MS}$Stack.

In recent decades, the field of multiobjective optimization has attracted considerable interest among evolutionary computation researchers. One of the main features that makes evolutionary methods particularly appealing for multiobjective problems is the implicit parallelism offered by a population, which enables simultaneous convergence toward the entire Pareto front. While a plethora of related algorithms have been proposed till date, a common attribute among them is that they focus on efficiently solving only a single optimization problem at a time. Despite the known power of implicit parallelism, seldom has an attempt been made to multitask, i.e., to solve multiple optimization problems simultaneously. It is contended that the notion of evolutionary multitasking leads to the possibility of automated transfer of information across different optimization exercises that may share underlying similarities, thereby facilitating improved convergence characteristics. In particular, the potential for automated transfer is deemed invaluable from the standpoint of engineering design exercises where manual knowledge adaptation and reuse are routine. Accordingly, in this paper, we present a realization of the evolutionary multitasking paradigm within the domain of multiobjective optimization. The efficacy of the associated evolutionary algorithm is demonstrated on some benchmark test functions as well as on a real-world manufacturing process design problem from the composites industry.

To enhance the search performance of evolutionary algorithms, reusing knowledge captured from past optimization experiences along the search process has been proposed in the literature, and demonstrated much promise. In the literature, there are generally three types of approaches for reusing knowledge from past search experiences, namely exact storage and reuse of past solutions, the reuse of model-based information, and the reuse of structured knowledge captured from past optimized solutions. In this paper, we focus on the third type of knowledge reuse for enhancing evolutionary search. In contrast to existing works, here we focus on knowledge transfer across heterogeneous continuous optimization problems with diverse properties, such as problem dimension, number of objectives, etc., that cannot be handled by existing approaches. In particular, we propose a novel autoencoding evolutionary search paradigm with learning capability across heterogeneous problems. The essential ingredient for learning structured knowledge from search experience in our proposed paradigm is a single layer denoising autoencoder (DA), which is able to build the connections between problem domains by treating past optimized solutions as the corrupted version of the solutions for the newly encountered problem. Further, as the derived DA holds a closed-form solution, the corresponding reusing of knowledge from past search experiences will not bring much additional computational burden on the evolutionary search. To evaluate the proposed search paradigm, comprehensive empirical studies on the complex multiobjective optimization problems are presented, along with a real-world case study from the fiber-reinforced polymer composites manufacturing industry.

In this paper, we present an evolutionary transfer reinforcement learning framework (eTL) for developing intelligent agents capable of adapting to the dynamic environment of multiagent systems (MASs). Specifically, we take inspiration from Darwin's theory of natural selection and Universal Darwinism as the principal driving forces that govern the evolutionary knowledge transfer process. The essential backbone of our proposed eTL comprises several meme-inspired evolutionary mechanisms, namely meme representation, meme expression, meme assimilation, meme internal evolution, and meme external evolution. Our proposed approach constructs social selection mechanisms that are modeled after the principles of human learning to identify appropriate interacting partners. eTL also models the intrinsic parallelism of natural evolution and errors that are introduced due to the physiological limits of the agents' ability to perceive differences, so as to generate “growth” and “variation” of knowledge that agents have of the world, thus exhibiting higher adaptivity capabilities on solving complex problems. To verify the efficacy of the proposed paradigm, comprehensive investigations of the proposed eTL against existing state-of-the-art TL methods in MAS, are conducted on the “minefield navigation tasks” platform and the “Unreal Tournament 2004” first person shooter computer game, in which homogeneous and heterogeneous learning machines are considered.

Meme automaton is an adaptive entity that autonomously acquires an increasing level of capability and intelligence through embedded memes evolving independently or via social interactions. This paper begins a study on memetic multiagent system (MeMAS) toward human-like social agents with memetic automaton. We introduce a potentially rich meme-inspired design and operational model, with Darwin's theory of natural selection and Dawkins' notion of a meme as the principal driving forces behind interactions among agents, whereby memes form the fundamental building blocks of the agents' mind universe. To improve the efficiency and scalability of MeMAS, we propose memetic agents with structured memes in this paper. Particularly, we focus on meme selection design where the commonly used elitist strategy is further improved by assimilating the notion of like-attracts-like in the human learning. We conduct experimental study on multiple problem domains and show the performance of the proposed MeMAS on human-like social behavior.

Traditionally, single-objective and multi-objective optimization have only considered a single problem in one run. However, the notion of evolutionary multitasking, which aims at solving multiple optimization problems simultaneously, has recently emerged in Evolutionary Computation (EC). It is inspired by the implicit parallelism of population-based search, which attempts to take advantage of implicit genetic transfer in a multitasking environment. According to optimization literature, transforming a single-objective optimization (SOO) problem into a multi-objective optimization (MOO) problem has often been found to remove local optima. Motivated by the aforementioned idea and the concept of multitasking, in this paper, we introduce a new strategy for tackling complex multi-modal problems. In particular, we solve the original (or target) SOO task together with an artificially formulated MOO task in a multitask setting. Therein, the MOO task is expected to provide a useful inductive bias to the search progress of the target SOO task by leveraging on the transferable knowledge shared between them, thereby helping overcome local optima and effectively guiding the population towards more promising regions of the search space.

Over the years, the algorithms of evolutionary computation have emerged as popular tools for tackling complex real-world optimization problems. A common feature among these algorithms is that they focus on efficiently solving a single problem at a time. Despite the availability of a population of individuals navigating the search space, and the implicit parallelism of their collective behavior, seldom has an effort been made to multitask. Considering the power of implicit parallelism, we are drawn to the idea that population-based search strategies provide an idyllic setting for leveraging the underlying synergies between objective function landscapes of seemingly distinct optimization tasks, particularly when they are solved together with a single population of evolving individuals. As has been recently demonstrated, allowing the principles of evolution to autonomously exploit the available synergies can often lead to accelerated convergence for otherwise complex optimization tasks. With the aim of providing deeper insight into the processes of evolutionary multitasking, we present in this paper a conceptualization of what, in our opinion, is one possible interpretation of the complementarity between optimization tasks. In particular, we propose a synergy metric that captures the correlation between objective function landscapes of distinct tasks placed in synthetic multitasking environments. In the long run, it is contended that the metric will serve as an important guide toward better understanding of evolutionary multitasking, thereby facilitating the design of improved multitasking engines.

Compression Resin Transfer Moulding is a popular method for high volume production of superior quality fibre-reinforced polymer composite parts. However, the process involves a large number of design variables that must be carefully chosen in order to reduce cycle time, capital layout and running costs, while maximizing final part quality. These objectives are principally governed by two separate phases of the manufacturing cycle, namely the resin filling and curing phases. It turns out that independently optimizing either phase (which is the general practice) may often lead to conditions that significantly restrict or even adversely affect the progress of the other. In light of this fact, a novel approach of modelling the entire composites manufacturing problem as bi-level program, one that assimilates both phases, has been adopted in this paper. In particular, an efficient multi-objective bi-level evolutionary algorithm is designed to effectively deal with the computationally expensive simulation-based optimization problem. The unique feature of the algorithm is that it incorporates a Pareto Rank Learning scheme, together with surrogate assistance for the upper level problem, in order to eliminate several expensive but redundant objective function evaluations. The optimization process is therefore considerably accelerated, assisting manufacturers in making improved decisions for this complex engineering design problem.

The Robust Vehicle Routing Problem with Time Windows has been gaining popularity over the past few years due to its focus on tackling uncertainty inherent to real world problems. Most of the current approaches in generating robust solutions require prior knowledge on the uncertainties, such as uncertainties in travel time. Hence, they are less than favorable to use in the absence of data, i.e., in the case of data starvation. In this paper, we present an evolutionary algorithm that in the absence of data on travel time uncertainty, provides a decision maker with a collection of solutions, each with a corresponding level of trade-off between total travel distance and solution robustness. In particular, we present a novel realization of route flexibility and its relation to solution robustness. Furthermore, we propose a bi-objective evolutionary algorithm for the vehicle routing problem with time windows where the objectives are (a) total travel distance and (b) solution flexibility. The proposed algorithm is tested on the well-known Solomon benchmarks and a trade-off analysis between total distance and solution flexibility is provided based on the obtained test results. Based on observations from the trade-off analysis, a number of suggestions to improve the current logistics system are provided.

Indicator-based evolutionary algorithm (IBEA1) is a fast and effective approach for solving multiobjective optimization problems (MOPs). In the classical IBEA1, the parameter κ is predefined to amplify or shrink the indicator differences on pairwise solutions. However, the value of κ in IBEA1 needs to be carefully calibrated based on the selected indicator (e.g., hypervolume or additive e-indicator) and the encountered MOPs. In this paper, a new version of IBEA1 (labeled as IBEA2 hereafter) is proposed to adaptively adjust parameter κ for solving various MOPs. The core idea of IBEA2 is to adapt parameter κ for the purpose of selecting the subset of offspring solutions with the maximum hypervolume into the next population. Experimental studies on 44 benchmark MOPs with 2-5 objectives in jMetal verified that IBEA2 is able to find higher hypervolumes against the four classical MOEAs, which are NSGAII, SPEA2, MOEA/D and IBEA1, in the literature.

The decomposition method in multiobjective evolutionary algorithms (MOEA/D) is an effective approach to evolve solutions along predefined weight vectors for solving multiobjective optimization problems (MOPs). However, obtaining evenly distributed weight vectors for different types of MOPs is a challenge problem especially when the true Pareto fronts (PFs) are unknown before a MOEA/D starts. In this paper, a new MOEA/D with a fast hypervolume archive (called FV-MOEA/D) is proposed to adaptively adjust the weight vectors for various shapes of PFs. The core idea of FV-MOEA/D is to periodically adjust weight vectors based on solutions in the proposed archive, in which convergence and diversity are maintained by maximizing hypervolume. Experimental studies on 58 benchmark MOPs in jMetal demonstrate that the proposed FV-MOEA/D not only reached higher hypervolumes when compare to five classical MOEAs i.e., NSGAII, SPEA2, IBEA, FV-MOEA and MOEA/D, but also obtained well distributed weight vectors on PFs with different geometrical characteristics.

Memetic algorithms (MAs), which have materialized as a fusion of population based global search and individual lifetime learning (i.e., local search) in the literature, have been widely used in real world applications to solve complex optimization problems. The balance of global and local search in MA plays a key role in defining the performance of MA in problem solving. The probabilistic memetic framework (PMF) was thus introduced to model MA as a process involving the decision of embracing the separate actions of global or local search. PMF balances these two actions by governing the local search intensity of each individual based on a theoretical upper bound derived while the search progresses. To use PMF for solving combinatorial optimization problems, according to our previous study [1], we note that the appropriate selection of a distance metric for estimating the local search intensity is a critical role. Nevertheless, to the best of our knowledge, little or no research works in the literature has studied on suitable distance metric for PMF in the context of combinatorial optimization problems. In this paper, we attempt to fill this gap by presenting a preliminary study on the selection of distance metric in PMF for capacitated arc routing problem (CARP). In particular, we first analyze the suitability of 4 existing popular distance metrics used in combinatorial optimization for solving CARP. Subsequently a score based on closeness of neighborhood and fitness landscape correlation is proposed to quantify the suitability of a distance metric in estimating the local search intensity for PMF in the context of combinatorial optimization. Experimental study on 24 egl CARP benchmark instances highlighted the significance of choice of appropriate distance metric in PMF for solving combinatorial optimization problems, with 4 new best known CARP solutions established in the present study.

In this paper, we present a multi-agent reinforcement learning (MARL) framework that leverages the emergent behaviors from swarm intelligence (SI). The essential backbone of our framework is an flocking-regulated cooperative learning paradigm in which the cooperation among learning agents is realized via the self-organizing principles derived from natural interaction of flocking boids. In the proposed MARL, each reinforcement learner learns and evolves in the dynamic environment, and is steered by flocking behavior rules such as cohesion, separation, alignment, fear, etcs. The use of the flocking rules provides distributed sensing and communication content for the cooperation of multiple learning agents in the context of pursuit game. The effectiveness of the MARL framework is studied by its application of the multi-agent pursuit game.

Memetic Multi-Agent System (MeMAS) has recently emerged as a combination of memetic automaton and multi-agent system (MAS), wherein all meme-inspired agents acquire increasing learning capacity and intelligence through meme evolution. This paper further presents a study of MeMAS in developing human-like non-player characters in complex first-person shooter (FPS) games. In particular, we consider a well-known commercial FPS game, known as Unreal Tournament 2004 (UT2004), as our game of interest and discuss the details of non-player characters based on a manifestation of “Temporal Difference - Fusion Architecture for Learning and Cognition” (TDFALCON) neural network. In addition, we present a brief cross-domain study of MeMAS wherein the useful knowledge in the form of memes learned from different yet related simple domains previously solved are used to enhance learning performance of the non-player characters in UT2004. Benchmark experiments are studied to investigate the efficacy of MeMAS in UT2004 from various aspects, including learning efficiency, generalization capability, and computational cost. The empirically results indicate that the MeMAS could clearly improve the learning effectiveness and efficiency of designed non-player characters in UT2004.

Auto-encoder is a popular representation learning technique which can capture the generative model of data via a encoding and decoding procedure typically driven by reconstruction errors in an unsupervised way. In this paper, we propose a semi-supervised manifold learning based auto-encoder (named semAE). semAE is based on a regularized auto-encoder framework which leverages semi-supervised manifold learning to impose regularization based on the encoded representation. Our proposed approach suits more practical scenarios in which a small number of labeled data are available in addition to a large number of unlabeled data. Experiments are conducted on several well-known benchmarking datasets to validate the efficacy of semAE from the aspects of both representation and classification. The comparisons to state-of-the-art representation learning methods on classification performance in semi-supervised settings demonstrate the superiority of our approach.

Today, modern databases with “Big Dimensionality” are experiencing a growing trend. Existing approaches that require the calculations of pairwise feature correlations in their algorithmic designs have scored miserably on such databases, since computing the full correlation matrix (i.e., square of dimensionality in size) is computationally very intensive (i.e., million features would translate to trillion correlations). This poses a notable challenge that has received much lesser attention in the field of machine learning and data mining research. Thus, this paper presents a study to fill in this gap. Our findings on several established databases with big dimensionality across a wide spectrum of domains have indicated that an extremely small portion of the feature pairs contributes significantly to the underlying interactions and there exists feature groups that are highly correlated. Inspired by the intriguing observations, we introduce a novel learning approach that exploits the presence of sparse correlations for the efficient identifications of informative and correlated feature groups from big dimensional data that translates to a reduction in complexity from O(m2n) to O(mlogm + Kamn), where Ka≪ min(m; n) generally holds. In particular, our proposed approach considers an explicit incorporation of linear and nonlinear correlation measures as constraints in the learning model. An efficient embedded feature selection strategy, designed to filter out the large number of non-contributing correlations that could otherwise confuse the classifier while identifying the correlated and informative feature groups, forms one of the highlights of our approach. We also demonstrated the proposed method on one-class learning, where notable speedup can be observed when solving one-class problem on big dimensional data. Further, to identify robust informative features with minimal sampling bias, our feature selection strategy embeds the V-fold cross validation in the learning model, so as to seek for features that exhibit stable or consistent performance accuracy on multiple data folds. Extensive empirical studies on both synthetic and several real-world datasets comprising up to 30 million dimensions are subsequently conducted to assess and showcase the efficacy of the proposed approach.

General game playing (GGP) aims at designing autonomous agents capable of playing any game within a certain genre, without human intervention. GGP agents accept the rules, which are written in the logic-based game definition language (GDL) and unknown to them beforehand, at runtime. The state-of-the-art players use Monte Carlo tree search (MCTS) together with the upper confidence bounds applied to trees (UCT) method. In this paper, we discuss several enhancements to GGP players geared towards more effective playing of single-player games within the MCTS/UCT framework. The main proposed improvements include introduction of a collection of lightweight policies which can be used for guiding the MCTS and a GGP-friendly way of using transposition tables. We have tested our base player and a specialized version of it for single-player games in a series of experiments using ten single-player games of various complexity. It is clear from the results that the optimized version of the player achieves significantly better performance. Furthermore, in the same set of tests against publicly available version of CadiaPlayer, one of the strongest GGP agents, the results are also favorable to the enhanced version of our player.

Computers and artificial intelligence play a key role in the production of artwork through the designing of synthetic agents that are able to reproduce the capabilities of human artists in assembling high-quality artefacts such as paintings and sculptures. In this context, music composition represents one of the art disciplines that can greatly benefit from the appropriate use of computational intelligence, as witnessed by the large number of research activities performed in this field over the recent years. Nevertheless, the automatic composition of music is far from being completely and precisely perfected due to the intrinsic virtuosity that characterizes human musicians' capabilities. This paper reduces this gap with the proposal of an intelligent scheme for the efficient composition of melodies based on a musical method that is inspired by and strongly characterized by human virtuosity: the unfigured bass technique. In particular, we formulate this music composition technique as an optimization problem and solve it with an adaptive multiagent memetic approach comprising diverse metaheuristics, the composer agents that cooperate to create high-quality four-voice pieces of music starting from a bass line as input. A collection of experimental studies on the famous Bach's four-voice chorales showed that the cooperation among different optimization strategies yields improved performance over the solutions obtained by conventional and hybrid evolutionary algorithms.

The design of evolutionary algorithms has typically been focused on efficiently solving a single optimization problem at a time. Despite the implicit parallelism of population-based search, no attempt has yet been made to multitask, i.e., to solve multiple optimization problems simultaneously using a single population of evolving individuals. Accordingly, this paper introduces evolutionary multitasking as a new paradigm in the field of optimization and evolutionary computation. We first formalize the concept of evolutionary multitasking and then propose an algorithm to handle such problems. The methodology is inspired by biocultural models of multifactorial inheritance, which explain the transmission of complex developmental traits to offspring through the interactions of genetic and cultural factors. Furthermore, we develop a cross-domain optimization platform that allows one to solve diverse problems concurrently. The numerical experiments reveal several potential advantages of implicit genetic transfer in a multitasking environment. Most notably, we discover that the creation and transfer of refined genetic material can often lead to accelerated convergence for a variety of complex optimization functions.

In this paper, a novel self-learning gene expression programming (GEP) methodology named SL-GEP is proposed to improve the search accuracy and efficiency of GEP. In contrast to the existing GEP variants, the proposed SL-GEP features a novel chromosome representation in which each chromosome is embedded with subfunctions that can be deployed to construct the final solution. As part of the chromosome, the subfunctions are self-learned or self-evolved by the proposed algorithm during the evolutionary search. By encompassing subfunctions or any partial solution as input arguments of another subfunction, the proposed SL-GEP facilitates the formation of sophisticated, higher-order, and constructive subfunctions that improve the accuracy and efficiency of the search. Further, a novel search mechanism based on differential evolution is proposed for the evolution of chromosomes in the SL-GEP. The proposed SL-GEP is simple, generic and has much fewer control parameters than the traditional GEP variants. The proposed SL-GEP is validated on 15 symbolic regression problems and six even-parity problems. Experimental results show that the proposed SL-GEP offers enhanced performances over several state-of-the-art algorithms in terms of accuracy and search efficiency.

This paper proposes a dynamic vehicle routing problem (DVRP) model with nonstationary stochastic travel times under traffic congestion. Depending on the traffic conditions, the travel time between two nodes, particularly in a city, may not be proportional to distance and changes both dynamically and stochastically over time. Considering this environment, we propose a Markov decision process model to solve this problem and adopt a rollout-based approach to the solution, using approximate dynamic programming to avoid the curse of dimensionality. We also investigate how to estimate the probability distribution of travel times of arcs which, reflecting reality, are considered to consist of multiple road segments. Experiments are conducted using a real-world problem faced by Singapore logistics/delivery company and authentic road traffic information.

Learning for maximizing AUC performance is an important research problem in machine learning. Unlike traditional batch learning methods for maximizing AUC which often suffer from poor scalability, recent years have witnessed some emerging studies that attempt to maximize AUC by single-pass online learning approaches. Despite their encouraging results reported, the existing online AUC maximization algorithms often adopt simple stochastic gradient descent approaches, which fail to exploit the geometry knowledge of the data observed in the online learning process, and thus could suffer from relatively slow convergence. To overcome the limitation of the existing studies, in this paper, we propose a novel algorithm of Adaptive Online AUC Maximization (AdaOAM), by applying an adaptive gradient method for exploiting the knowledge of historical gradients to perform more informative online learning. The new adaptive updating strategy by AdaOAM is less sensitive to parameter settings due to its natural effect of tuning the learning rate. In addition, the time complexity of the new algorithm remains the same as the previous non-adaptive algorithms. To demonstrate the effectiveness of the proposed algorithm, we analyze its theoretical bound, and further evaluate its empirical performance on both public benchmark datasets and anomaly detection datasets. The encouraging empirical results clearly show the effectiveness and efficiency of the proposed algorithm.

Bilevel optimization is a type of mathematical program in which one optimization problem (called the lower level problem) is nested within another (called the upper level problem). In recent years, there has been considerable interest in the development of algorithms that can handle multiple objective functions at both levels. The challenge lies in the implication that every upper level decision leads to a set of Pareto optimal solutions at the lower level. As a result, a single point in the upper level decision space maps to a set of points in the upper level objective space. Since standard multiobjective evolutionary algorithms (MOEAs) are not designed for such point-to-set mappings, the state-of-the-art solution methods often dictate several enhancements to existing MOEAs. In this paper, we propose an adaptive scalarization based approach to solving bilevel programs with multiple objectives at both levels, such that any off-the-shelf MOEA can be used with minimum modification. Subsequently, we put forward a surrogate-assistance technique that can significantly lower the computational cost commonly associated with such problems. Finally, proof-of-concept numerical experiments are carried out in order to demonstrate the potential of our proposed methods.

Profitable tour problem (PTP) belongs to the class of vehicle routing problem (VRP) with profits seeking to maximize the difference between the total collected profit and the total cost incurred. Traditionally, PTP involves single vehicle. In this paper, we consider PTP with multiple vehicles. Unlike the classical VRP that seeks to serve all customers, PTP involves the strategic-level customer selection so as to maximize the total collected profit and the operational-level route optimization to minimize the total cost incurred. Therefore, PTP is essentially the knapsack problem at the strategic level with VRP at the operational level. That means the evolutionary bi-level programming would be a suitable choice of methodology for solving the NP-hard PTP. Employing some evolutionary method to solve the bi-level program naively would undoubtedly be prohibitively expensive. We thus present in this paper the notion of knowledge adoption to approximate the initial solution to the lower-level optimization problem for a given trial solution of the upper-level decision variables. One may consider the knowledge adoption as a special case of knowledge transfer in which the transfer takes place within the same problem domain. Refining the approximate initial solution with local search forces it to quickly converge to some locally optimal solution. The better the estimation of the initial solution, the closer the local optimum will be to the global one. PTP finds its important application in the fields of transportation and logistics. In addressing last-mile problem using auction at the urban consolidation center (UCC), PTP plays a significant role in the winner determination problem (WDP) that follows. Empirical study demonstrates the efficacy of the proposed approach in solving the PTP-based WDP, yielding significantly higher profit, utilization, and service level compared to when the UCC adopts the conventional WDP based on multiple knapsack problem (i.e. the MKP-based WDP).

To find diversified solutions converging to true Pareto fronts (PFs), hypervolume (HV) indicator-based algorithms have been established as effective approaches in multiobjective evolutionary algorithms (MOEAs). However, the bottleneck of HV indicator-based MOEAs is the high time complexity for measuring the exact HV contributions of different solutions. To cope with this problem, in this paper, a simple and fast hypervolume indicator-based MOEA (FV-MOEA) is proposed to quickly update the exact HV contributions of different solutions. The core idea of FV-MOEA is that the HV contribution of a solution is only associated with partial solutions rather than the whole solution set. Thus, the time cost of FV-MOEA can be greatly reduced by deleting irrelevant solutions. Experimental studies on 44 benchmark multiobjective optimization problems with 2-5 objectives in platform jMetal demonstrate that FV-MOEA not only reports higher hypervolumes than the five classical MOEAs (nondominated sorting genetic algorithm II (NSGAII), strength Pareto evolutionary algorithm 2 (SPEA2), multiobjective evolutionary algorithm based on decomposition (MOEA/D), indicator-based evolutionary algorithm, and S-metric selection based evolutionary multiobjective optimization algorithm (SMS-EMOA)), but also obtains significant speedup compared to other HV indicator-based MOEAs.

In recent decades, a plethora of dedicated evolutionary algorithms (EAs) have been crafted to solve domain-specific complex problems more efficiently. Many advanced EAs have relied on the incorporation of domain-specific knowledge as inductive biases that is deemed to fit the problem of interest well. As such, the embedment of domain knowledge about the underlying problem within the search algorithms is becoming an established mode of enhancing evolutionary search performance. In this paper, we present a study on evolutionary memetic computing paradigm that is capable of learning and evolving knowledge meme that traverses different but related problem domains, for greater search efficiency. Focusing on combinatorial optimization as the area of study, a realization of the proposed approach is investigated on two NP-hard problem domains (i.e., capacitated vehicle routing problem and capacitated arc routing problem). Empirical studies on well-established routing problems and their respective state-of-the-art optimization solvers are presented to study the potential benefits of leveraging knowledge memes that are learned from different but related problem domains on future evolutionary search.

Lately, the Vehicle Routing Problem (VRP) in the city, known as City VRP, has gained popularity with its importance in city logistics. Similar to city logistics, City VRP mainly differs from conventional VRP in terms of the stakeholders involved, namely the shipper, carrier, resident, and administrator. Accordingly, this paper surveys the City VRP literature categorized by stakeholders and summarizes the constraints, models, and solution methods for VRP in urban cities. City VRPs are also analyzed based on the problem of interest considered by the stakeholders and the corresponding models that have been proposed in response. Through this review, we identify the state of the art of City VRP, highlight the core challenging issues, and suggest some potential research area in this field that have remained underexplored.

Standard Regression models are presented with n samples from an input space X that is composed of observational data of the form (xi, y(xi)), i = 1...n where each xidenotes a k-dimensional input vector of design variables and y is the response. When k ≫ n, high variance and over-fitting become a major concern. In this paper we propose a novel approach to mitigate this problem by transforming the input vectors into new smaller vectors (called Z set) using only a set of simple statistical moments. Genetic Algorithm (GA) has been used to evolve a transformation procedure. It is used to optimise an optimal sequence of statistical moments and their input parameters. We used Linear Regression (LR) as an example to quantify the quality of the evolved transformation procedure. Empirical evidences, collected from benchmark functions and real-world problems, demonstrate that the proposed transformation approach is able to dramatically improve LR generalisation and make it outperform other state-of-the-art regression models such as Genetic Programming, Kriging, and Radial Basis Functions Networks. In addition, we present an analysis to shed light on the most important statistical moments that are useful for the transformation process.

This paper proposes a methodology to create an interpretable fuzzy model for monthly rainfall time series prediction. The proposed methodology incorporates the advantages of artificial neural network, fuzzy logic and genetic algorithm. In the first step, the differences between the time series data are calculated and they are used to define the interval between the membership functions of a Mamdani-type fuzzy inference system. Next, artificial neural network is used to develop the model from input-output data and the established model is then used to extract the fuzzy rules. The parameters of the created fuzzy model are then optimized by using genetic algorithm. The proposed model was applied to eight monthly rainfall time series data in the northeast region of Thailand. The experimental results showed that the proposed model provided satisfactory prediction accuracy when compared to other commonly-used prediction models. Due to the interpretability nature of the model, human analysts can gain insight knowledge of the data to be modeled.

An important consideration of multiobjective optimization (MOO) is the quantitative metrics used for defining the optimality of different solution sets, which is also the basic principle for the design and evaluation of MOO algorithms. Although a plethora of performance metrics have been proposed in the MOO context, there has been a lack of insights on the relationships between metrics. In this paper, we first group the major MOO metrics proposed to date according to four core performance criteria considered in the literature, namely, capacity, convergence, diversity, and convergence-diversity. Then, a comprehensive study is conducted to investigate the relationships among representative group metrics, including generational distance, ϵ-indicator (I(1)ϵ+), spread (∆), generalized spread (∆∗), inverted generational distance, and hypervolume. Experimental results indicated that these six metrics show high consistencies when Pareto fronts (PFs) are convex, whereas they show certain contradictions on concave PFs.

Memetic multiagent system emerges as an enhanced version of multiagent systems with the implementation of meme-inspired computational agents. It aims to evolve human-like behavior of multiple agents by exploiting the Dawkins' notion of a meme and Universal Darwinism. Previous research has developed a computational framework in which a series of memetic operations have been designed for implementing human-like agents. This paper will focus on improving the human-like behavior of multiple agents when they are engaged in social interactions. The improvement is mainly on how an agent shall learn from others and adapt its behavior in a complex dynamic environment. In particular, we design a new mechanism that supervises how the agent shall select one of the other agents for the learning purpose. The selection is a trade-off between the elitist and like-attracts-like principles. We demonstrate the desirable interactions of multiple agents in two problem domains.

In this paper we present a negotiation agent based on Genetic Algorithm (GA) and Surrogate Modelling for a multi-player multi-issue negotiation model under incomplete information scenarios to solve a resource-allocation problem. We consider a multi-lateral negotiation protocol by which agents make offers sequentially in consecutive rounds until the deadline is reached. Agents' offers represent suggestions about how to divide the available resources among all agents participating in the negotiation. Each agent may “Accept” or “Reject” the offers made by its opponents through selecting the “Accept” or “Reject” option. The GA is used to explore the space of offers and surrogates used to model the behaviours of individual opponent agents for enhanced genetic evolution of offers that is agreeable upon all agents. The GA population comprises of solution individuals that are formulated as matrices where a specialised three different search operators that take the matrix representation into considerations are considered. Experimental studies of the proposed negotiation agent under different scenarios demonstrated that the negotiations by the agents completed in agreement before the deadline is reached, while at the same time, maximising profits.

Due to the complex nature and numerous interacting factors that contributes to the increased vulnerability of supply chains, traditional methods have been found to be inadequate for Supply Chain Risk Management (SCRM). Agent-Based Modeling and Simulation (ABMS), an agent-oriented approach to model and simulate complex adaptive systems, represents a recent development in supply chain planning that has been regarded highly appropriate for studying risk management [1-4]. The objective of this paper is to provide a multi-perspective survey of the state-of-the-art agent based modeling and simulation approaches for SCRM.

Supply chain disruptions and risks have been garnering more attention in recent years owing to the occurrence of several high-profile disruptions and many published works acknowledge the increasing vulnerability to disruptions in supply chains. Few, however, explicitly embrace complex systems (CS) approach as a holistic, system-wide perspective towards supply chain risk management (SCRM). This paper analyzes the current status and identifies research trends in applying CS approaches to SCRM by analyzing the Science Direct publication database. The paper then provides a further review of CS technologies for SCRM and discusses on the gaps and directions for future research in SCRM by CS technologies.

The paper presents an agent-based simulation to model supply chains (SCs) and evaluate the bullwhip effect, an important index to measure the stability of SCs, under the stochastic demand and lead time. We carried out a series of numerical experiments to measure the bull whip effects in two scenarios. In the first scenario with a simple 2-tier SC, the results of simulation were obtained and compared directly to the analytical counterparts in literature. From the comparison between simulated and analytical results, it was figured out that the assumption of non-negative demand/order quantities, which are usually not considered in analytical models, could affect the results of bullwhip effect measurement. In the second scenario, the bullwhip effect in a 4-tier SC was evaluated to observe quantitatively how the variances in demands and replenishment orders are amplified while moving upwards the SC.

To deal with complex optimization problems plagued with computationally expensive fitness functions, the use of surrogates to replace the original functions within the evolutionary framework is becoming a common practice. However, the appropriate datacentric approximation methodology to use for the construction of surrogate model would depend largely on the nature of the problem of interest, which varies from fitness landscape and state of the evolutionary search, to the characteristics of search algorithm used. This has given rise to the plethora of surrogate-assisted evolutionary frameworks proposed in the literature with ad hoc approximation/surrogate modeling methodologies considered. Since prior knowledge on the suitability of the data centric approximation methodology to use in surrogate-assisted evolutionary optimization is typically unavailable beforehand, this paper presents a novel evolutionary framework with the evolvability learning of surrogates (EvoLS) operating on multiple diverse approximation methodologies in the search. Further, in contrast to the common use of fitness prediction error as a criterion for the selection of surrogates, the concept of evolvability to indicate the productivity or suitability of an approximation methodology that brings about fitness improvement in the evolutionary search is introduced as the basis for adaptation. The backbone of the proposed EvoLS is a statistical learning scheme to determine the evolvability of each approximation methodology while the search progresses online. For each individual solution, the most productive approximation methodology is inferred, that is, the method with highest evolvability measure. Fitness improving surrogates are subsequently constructed for use within a trust-region enabled local search strategy, leading to the self-configuration of a surrogate-assisted memetic algorithm for solving computationally expensive problems. A numerical study of EvoLS on commonly used benchmark problems and a real-world computationally expensive aerodynamic car rear design problem highlights the efficacy of the proposed EvoLS in attaining reliable, high quality, and efficient performance under a limited computational budget.

Domain adaptation (DA), which leverages labeled data from related source domains, comes in handy when the label information of the target domain is scarce or unavailable. However, as the source data do not come from the same origin as that of the target domain, the predictive distributions of the source and target domains are likely to differ in reality. At the extreme, the predictive distributions of the source domains can differ completely from that of the target domain. In such case, using the learned source classifier to assist in the prediction of target data can result in prediction performance that is poorer than that with the omission of the source data. This phenomenon is established as negative transfer with impact known to be more severe in the multiclass context. To combat negative transfer due to differing predictive distributions across domains, we first introduce the notion of positive transferability for the assessment of synergy between the source and target domains in their prediction models, and we also propose a criterion to measure the positive transferability between sample pairs of different domains in terms of their prediction distributions. With the new measure, a predictive distribution matching (PDM) regularizer and a PDM framework learn the target classifier by favoring source data with large positive transferability while inferring the labels of target unlabeled data. Extensive experiments are conducted to validate the performance efficacy of the proposed PDM framework using several commonly used multidomain benchmark data sets, including Sentiment, Reuters, and Newsgroup, in the context of both binary-class and multiclass domains. Subsequently, the PDM framework is put to work on a real-world scenario pertaining to water cluster molecule identification. The experimental results illustrate the adverse impact of negative transfer on several state-of-the-art DA methods, whereas the proposed framework exhibits excellent and robust predictive performances.

Designing a classifier in the absence of labeled data is becoming a common encounter as the acquisition of informative labels is often difficult or expensive, particularly on new uncharted target domains. The feasibility of attaining a reliable classifier for the task of interest is embarked by some in transfer learning, where label information from relevant source domains is considered for complimenting the design process. The core challenge arising from such endeavors, however, is the induction of source sample selection bias, such that the trained classifier has the tendency of steering toward the distribution of the source domain. In addition, this bias is deemed to become more severe on data involving multiple classes. Considering this cue, our interest in this paper is to address such a challenge in the target domain, where ordinal labeled data are unavailable. In contrast to the previous works, we propose a transfer ordinal label learning paradigm to predict the ordinal labels of target unlabeled data by spanning the feasible solution space with ensemble of ordinal classifiers from the multiple relevant source domains. Specifically, the maximum margin criterion is considered here for the construction of the target classifier from an ensemble of source ordinal classifiers. Theoretical analysis and extensive empirical studies on real-world data sets are presented to study the benefits of the proposed method.

A major drawback of evolutionary optimization approaches in the literature is the apparent lack of automated knowledge transfers and reuse across problems. Particularly, evolutionary optimization methods generally start a search from scratch or ground zero state, independent of how similar the given new problem of interest is to those optimized previously. In this paper, we present a study on the transfer of knowledge in the form of useful structured knowledge or latent patterns that are captured from previous experiences of problem-solving to enhance future evolutionary search. The essential contributions of our present study include the meme learning and meme selection processes. In contrast to existing methods, which directly store and reuse specific problem solutions or problem sub-components, the proposed approach models the structured knowledge of the strategy behind solving problems belonging to similar domain, i.e., via learning the mapping from problem to its corresponding solution, which is encoded in the form of identified knowledge representation. In this manner, knowledge transfer can be conducted across problems, from differing problem size, structure to representation, etc. A demonstrating case study on the capacitated arc routing problem (CARP) is presented. Experiments on benchmark instances of CARP verified the effectiveness of the proposed new paradigm.

Non-photorealistic rendering systems strive to create compelling stylized effects from realistic images. We present an interactive process using flocks of autonomous agents to model a painter's brush. As flocks of agents glide across the canvas like bristles on a paint brush, a stylized picture can be produced by carefully directing the path of movement. The agents leave behind a trail of color resulting in painterly or pencil sketch looking images.

In this paper, we present a novel cross-surrogate assisted memetic algorithm (CSAMA) as a manifestation of multi co-objective evolutionary computation to enhance the search on computationally expensive problems by means of transferring, sharing and reusing information across objectives. In particular, the construction of surrogate for one objective is augmented with information from other related objectives to improve the prediction quality. The process is termed as a cross-surrogate modelling methodology, which will be used in lieu with the original expensive functions during the evolutionary search. Analyses on the prediction quality of the cross-surrogate modelling and the search performance of the proposed algorithm are conducted on the benchmark problems with assessments made against several state-of-the-art multiobjective evolutionary algorithms. The results obtained highlight the efficacy of the proposed CSAMA in attaining high quality Pareto optimal solutions under limited computational budget.

In this paper, the interest is on cases where assessing the goodness of a solution for the problem is costly or hazardous to construct or extremely computationally intensive to compute. We label such category of problems as “expensive” in the present study. In the context of multi-objective evolutionary optimizations, the challenge amplifies, since multiple criteria assessments, each defined by an “expensive” objective is necessary and it is desirable to obtain the Pareto-optimal solution set under a limited resource budget. To address this issue, we propose a Pareto Rank Learning scheme that predicts the Pareto front rank of the offspring in MOEAs, in place of the “expensive” objectives when assessing the population of solutions. Experimental study on 19 standard multi-objective benchmark test problems concludes that Pareto rank learning enhanced MOEA led to significant speedup over the state-of-the-art NSGA-II, MOEA/D and SPEA2.

Evolutionary data, such as topic changing blogs and evolving trading behaviors in capital market, is widely seen in business and social applications. The time factor and intrinsic change embedded in evolutionary data greatly challenge evolutionary clustering. To incorporate the time factor, existing methods mainly regard the evolutionary clustering problem as a linear combination of snapshot cost and temporal cost, and reflect the time factor through the temporal cost. It still faces accuracy and scalability challenge though promising results gotten. This paper proposes a novel evolutionary clustering approach, evolutionary maximum margin clustering (e-MMC), to cluster large-scale evolutionary data from the maximum margin perspective. e-MMC incorporates two frameworks: Data Integration from the data changing perspective and Model Integration corresponding to model adjustment to tackle the time factor and change, with an adaptive label allocation mechanism. Three e-MMC clustering algorithms are proposed based on the two frameworks. Extensive experiments are performed on synthetic data, UCI data and real-world blog data, which confirm that e-MMC outperforms the state-of-the-art clustering algorithms in terms of accuracy, computational cost and scalability. It shows that e-MMC is particularly suitable for clustering large-scale evolving data.

Modern GPUs enable widely affordable personal computers to carry out massively parallel computation tasks. NVIDIA's CUDA technology provides a wieldy parallel computing platform. Many state-of-the-art algorithms arising from different fields have been redesigned based on CUDA to achieve computational speedup. Differential evolution (DE), as a very promising evolutionary algorithm, is highly suitable for parallelization owing to its data-parallel algorithmic structure. However, most existing CUDA-based DE implementations suffer from excessive low-throughput memory access and less efficient device utilization. This work presents an improved CUDA-based DE to optimize memory and device utilization: several logically-related kernels are combined into one composite kernel to reduce global memory access; kernel execution configuration parameters are automatically determined to maximize device occupancy; streams are employed to enable concurrent kernel execution to maximize device utilization. Experimental results on several numerical problems demonstrate superior computational time efficiency of the proposed method over two recent CUDA-based DE and the sequential DE across varying problem dimensions and algorithmic population sizes.

In the absence of the labeled samples in a domain referred to as target domain, Domain Adaptation (DA) techniques come in handy. Generally, DA techniques assume there are available source domains that share similar predictive function with the target domain. Two core challenges of DA typically arise, variance that exists between source and target domains, and the inherent source hypothesis bias. In this paper, we first propose a Stability Transfer criterion for selecting relevant source domains with low variance. With this criterion, we introduce a TARget learning Assisted by Source Classifier Adaptation (TARASCA) method to address the two core challenges that have impeded the performances of DA techniques. To verify the robustness of TARASCA, extensive experimental studies are carried out with comparison to several state-of-the-art DA methods on the real-world Sentiment and Newsgroups datasets, where various settings for the class ratios of the source and target domains are considered.

Predicting binding between macromolecule and small molecule is a crucial phase in the field of rational drug design. AutoDock Vina, one of the most widely used docking software released in 2009, uses an empirical scoring function to evaluate the binding affinity between the molecules and employs the iterated local search global optimizer for global optimization, achieving a significantly improved speed and better accuracy of the binding mode prediction compared its predecessor, AutoDock 4. In this paper, we propose further improvement in the local search algorithm of Vina by heuristically preventing some intermediate points from undergoing local search. Our improved version of Vina-dubbed QVina-achieved a maximum acceleration of about 25 times with the average speed-up of 8.34 times compared to the original Vina when tested on a set of 231 protein-ligand complexes while maintaining the optimal scores mostly identical. Using our heuristics, larger number of different ligands can be quickly screened against a given receptor within the same time frame.

Ordinal regression is commonly formulated as a multiclass problem with ordinal constraints. The challenge of designing accurate classifiers for ordinal regression generally increases with the number of classes involved, due to the large number of labeled patterns that are needed. The availability of ordinal class labels, however, is often costly to calibrate or difficult to obtain. Unlabeled patterns, on the other hand, often exist in much greater abundance and are freely available. To take benefits from the abundance of unlabeled patterns, we present a novel transductive learning paradigm for ordinal regression in this paper, namely transductive ordinal regression (TOR). The key challenge of this paper lies in the precise estimation of both the ordinal class label of the unlabeled data and the decision functions of the ordinal classes, simultaneously. The core elements of the proposed TOR include an objective function that caters to several commonly used loss functions casted in transductive settings, for general ordinal regression. A label swapping scheme that facilitates a strictly monotonic decrease in the objective function value is also introduced. Extensive numerical studies on commonly used benchmark datasets including the real-world sentiment prediction problem are then presented to showcase the characteristics and efficacies of the proposed TOR. Further, comparisons to recent state-of-the-art ordinal regression methods demonstrate the introduced transductive learning paradigm for ordinal regression led to the robust and improved performance.

In science, gene provides the instruction for making proteins, while meme is the sociocultural equivalent of a gene containing instructions for carrying out behavior. Taking inspiration from nature, we model the memeplex in search as instructions that specify the coadapted meme complexes of individuals in their lifetime. In particular, this paper presents a study on the conceptual modeling of meme complexes or memeplexes for more effective problem solving in the context of modern stochastic optimization. The memeplex representation, credit assignment criteria for meme coadaptation, and the role of emergent memeplexes in the lifetime learning process of a memetic algorithm in search are presented. A coadapted memetic algorithm that takes the proposed conceptual modeling of memeplexes into actions to solve capacitated vehicle routing problems (CVRPs) of diverse characteristics is then designed. Results showed that adaptive memeplexes provide a means of creating highly robust, self-configuring, and scalable algorithms, thus generating improved or competitive results when benchmarking against several existing adaptive or human-designed state-of-the-art memetic algorithms and metaheuristics, on a plethora of CVRP sets considered.

Memetics is a new science that has attracted in creasing attentions in the recent decades. Beyond the formalism of simple hybrids, adaptive hybrids and memetic algorithms, the notion of memetic automaton as an adaptive entity that is self-contained and uses memes as building blocks of information is recently conceptualized in the context of computational intelligence as potential tools for effective problem-solving. Taking this cue, this paper embarks a study on Memetic Multi agent system (MeM) towards human-like social agents with memetic automaton. Particularly, we introduce a potentially rich meme-inspired design and operational model, with Darwin's theory of natural selections and Dawkins' notion of a meme as the principal driving forces behind interactions among agents, whereby memes formed the fundamental building blocks of the agents' mind universe. Experimental studies on a Mine Navigation Task indicates the modeling of memetic agents that resemble the natural way of human interaction can lead to greater level of adaptivity and effective problem-solving.

Stochastic optimization of computationally expensive problems is a relatively new field of research in evolutionary computation (EC). At present, few EC works have been published to handle problems plagued with constraints that are expensive to compute. This paper presents a surrogate-assisted memetic co-evolutionary framework to tackle both facets of practical problems, i.e. the optimization problems having computationally expensive objectives and constraints. In contrast to existing works, the cooperative co-evolutionary mechanism is adopted as the backbone of the framework to improve the efficiency of surrogate-assisted evolutionary techniques. The idea of random-problem decomposition is introduced to handle interdependencies between variables, eliminating the need to determine the decomposition in an ad-hoc manner. Further, a novel multi-objective ranking strategy of constraints is also proposed. Empirical results are presented for a series of commonly used benchmark problems to validate the proposed algorithm.

The success of Memetic Algorithms (MAs) has driven many researchers to be more focused on the efficiency aspect of the algorithms such that it would be possible to effectively employ MAs to solve computationally expensive optimization problems where single evaluation of the objective and constraint functions may require minutes to hours of CPU time. One of the important design issues in MAs is the choice of the individuals upon which local search procedure should be applied. Selecting only some potential individuals lessens the demand for functional evaluations hence accelerates convergence to the global optimum. In recent years, advances have been made targeting optimization problems with single equality constraint h(x) = 0. The presence of previously evaluated candidate solutions with different signs of constraint values within some localities thus allows the estimation of the constraint boundary. An individual will undergo local search only if it is sufficiently close to the approximated boundary. Elegant as it may seem, the approach had unfortunately assumed that every constraint function maps the design variables to optimize into unbounded real values. This, however, may not always be the case in practice. In this paper, we present a strategy to efficiently solve constrained problems with a single equality constraint; the function of which maps the design variables into restricted (either strictly non-negative or strictly non-positive) real values only.

Domain Adaptation (DA) methods are usually carried out by means of simply reducing the marginal distribution differences between the source and target domains, and subsequently using the resultant trained classifier, namely source classifier, for use in the target domain. However, in many cases, the true predictive distributions of the source and target domains can be vastly different especially when their class distributions are skewed, causing the issues of sample selection bias in DA. Hence, DA methods which leverage the source labeled data may suffer from poor generalization in the target domain, resulting in negative transfer. In addition, we observed that many DA methods use either a source classifier or a linear combination of source classifiers with a fixed weighting for predicting the target unlabeled data. Essentially, the labels of the target unlabeled data are spanned by the prediction of these source classifiers. Motivated by these observations, in this paper, we propose to construct many source classifiers of diverse biases and learn the weight for each source classifier by directly minimizing the structural risk defined on the target unlabeled data so as to heal the possible sample selection bias. Since the weights are learned by maximizing the margin of separation between opposite classes on the target unlabeled data, the proposed method is established here as Maximal Margin Target Label Learning (MMTLL), which is in a form of Multiple Kernel Learning problem with many label kernels. Extensive experimental studies of MMTLL against several state-of-the-art methods on the Sentiment and Newsgroups datasets with various imbalanced class settings showed that MMTLL exhibited robust accuracies on all the settings considered and was resilient to negative transfer, in contrast to other counterpart methods which suffered significantly in prediction accuracy.

RELIEF is a very effective and extremely popular feature selection algorithm developed for the first time in 1992 by Kira and Rendell. Since then it has been modified and expanded in various ways to make it more efficient. But the original RELIEF and all of its expansions are for feature selection over labeled data for classification purposes. To the best of our knowledge, for the first time ever RELIEF is used in this paper as RELIEF-C for unlabeled data to select relevant features for clustering. We modified RELIEF so as to overcome its inherent difficulties in the presence of large number of irrelevant features and/or significant number of noisy tuples. RELIEF-C has several advantages over existing wrapper and filter feature selection methods: (a) it works well in the presence of large amount of noisy tuples, (b) it is robust even when underlying clustering algorithm fails to cluster properly, and (c) it accurately recognizes the relevant features even in the presence of large number of irrelevant features. We compared RELIEF-C with two established feature selection methods for clustering. RELIEF-C outperforms other methods significantly over synthetic, benchmark and real world data sets particularly when data set consists of large amount of noisy tuples and/or irrelevant features.

The importance of domain knowledge in the design of effective evolutionary algorithms (EAs) is widely acknowledged in the meta-heuristics community. In the last few decades, a plethora of EAs has been manually designed by domain experts for solving domain-specific problems. Specialization has been achieved mainly by embedding available domain knowledge into the algorithms. Although programming libraries have been made available to construct EAs, a unifying framework for designing specialized EAs across different problem domains and branches of evolutionary computing does not exist yet. In this paper, we address this issue by introducing an evolutionary computing modeling language (ECML) which is based on the unified modeling language (UML). ECML incorporates basic UML elements and introduces new extensions that are specially needed for the evolutionary computation domain. Subsequently, the concept of meta evolutionary algorithms (MEAs) is introduced as a family of EAs that is capable of interpreting ECML. MEAs are solvers that are not restricted to a particular problem domain or branch of evolutionary computing through the use of ECML. By separating problem-specific domain knowledge from the EA implementation, we show that a unified framework for evolutionary computation can be attained. We demonstrate our approach by applying it to a number of examples.

Memetic computation is a paradigm that uses the notion of meme(s) as units of information encoded in computational representations for the purpose of problem-solving. It covers a plethora of potentially rich meme-inspired computing methodologies, frameworks and operational algorithms including simple hybrids, adaptive hybrids and memetic automaton. In this paper, a comprehensive multi-facet survey of recent research in memetic computation is presented.

The five papers in this special issue focus on memetic computation.

Capacitated arc routing problem (CARP) has attracted much attention due to its generality to many real world problems. Memetic algorithm (MA), among other meta-heuristic search methods, has been shown to achieve competitive performances in solving CARP ranging from small to medium size. In this paper we propose a formal probabilistic memetic algorithm for CARP that is equipped with an adaptation mechanism to control the degree of global exploration against local exploitation while the search progresses. Experimental study on benchmark instances of CARP showed that the proposed probabilistic scheme led to improved search performances when introduced into a recently proposed state-of-the-art MA. The results obtained on 24 instances of the capacitated arc routing problems highlighted the efficacy of the probabilistic scheme with 9 new best known solutions established.

In orthodontics, retraction springs made of metallic wires are often used to move a tooth with respect to another by the virtue of the spring back effect. Specially selected form of spring may result in accurate force and moment required to move the tooth towards direction that suits a particular patient. In current practice, the geometry remains to be selected manually by orthodontists and no substantial automation of such process has been proposed to date. In this paper, we experiment with the automated geometry selection of the orthodontic retraction spring using constrained evolutionary optimization. Particularly, a Classifier-assisted Constrained Memetic Algorithm (CCMA) is designed for the purpose. The main feature of CCMA lies in the ability to identify appropriate spring structures that should undergo further refinement using a classifier system to perform the inference. Comparison to the baseline canonical Genetic Algorithm (GA) and Memetic Algorithm (MA) further highlights the efficacy of the proposed approach. In addition, to also assert the robustness of the CCMA for general complex design, further studies on commonly used constrained benchmark problems and existing constrained evolutionary optimization methods are also reported in the paper.

We have designed and implemented a 3D virtual world based on the Co-Space concept encompasses the Youth Olympic Village (YOV) and several sports competition venues. It is a massively multiplayer online (MMO) virtual world built according to the actual, physical locations of the YOV and sports competition venues. On top of that, the Co-Space is being populated with human-like avatars, which are created according to the actual human size and appearance, they perform their activities and interact with the users in real-world context. In addition, autonomous intelligent agents are integrated into the Co-Space to provide context-aware and personalized services to the users. These agents, which are in the form of human-like avatars, act as the users' personal tour guides, they interact with the users and gradually learn the users' preferences. Then, the agents are able to suggest suitable locations for the users to visit. The intention of deploying these agents is to enrich the Co-Space with a variety of interactions tailored to each user's preferences instead of the same contents for everyone. Besides, our Co-Space also provides an avenue for people from different locations throughout the world to explore the YOV and experience the atmosphere of the sports events in Singapore.

Multiclass cancer classification on microarray data has provided the feasibility of cancer diagnosis across all of the common malignancies in parallel. Using multiclass cancer feature selection approaches, it is now possible to identify genes relevant to a set of cancer types. However, besides identifying the relevant genes for the set of all cancer types, it is deemed to be more informative to biologists if the relevance of each gene to specific cancer or subset of cancer types could be revealed or pinpointed. In this paper, we introduce two new definitions of multiclass relevancy features, i.e., full class relevant (FCR) and partial class relevant (PCR) features. Particularly, FCR denotes genes that serve as candidate biomarkers for discriminating all cancer types. PCR, on the other hand, are genes that distinguish subsets of cancer types. Subsequently, a Markov blanket embedded memetic algorithm is proposed for the simultaneous identification of both FCR and PCR genes. Results obtained on commonly used synthetic and real-world microarray data sets show that the proposed approach converges to valid FCR and PCR genes that would assist biologists in their research work. The identification of both FCR and PCR genes is found to generate improvement in classification accuracy on many microarray data sets. Further comparison study to existing state-of-the-art feature selection algorithms also reveals the effectiveness and efficiency of the proposed approach.

An important issue in designing memetic algorithms (MAs) is the choice of solutions in the population for local refinements, which becomes particularly crucial when solving computationally expensive problems. With single evaluation of the objective/constraint functions necessitating tremendous computational power and time, it is highly desirable to be able to focus search efforts on the regions where the global optimum is potentially located so as not to waste too many function evaluations. For constrained optimization, the global optimum must either be located at the trough of some feasible basin or some particular point along the feasibility boundary. Presented in this paper is an instance of optinformatics where a new concept of modeling the feasibility structure of inequality-constrained optimization problems-dubbed the feasibility structure modeling-is proposed to perform geometrical predictions of the locations of candidate solutions in the solution space: deep inside any infeasible region, nearby any feasibility boundary, or deep inside any feasible region. This knowledge may be unknown prior to executing an MA but it can be mined as the search for the global optimum progresses. As more solutions are generated and subsequently stored in the database, the feasibility structure can thus be approximated more accurately. As an integral part, a new paradigm of incorporating the classification-rather than the regression-into the framework of MAs is introduced, allowing the MAs to estimate the feasibility boundary such that effective assessments of whether or not the candidate solutions should experience local refinements can be made. This eventually helps preventing the unnecessary refinements and consequently reducing the number of function evaluations required to reach the global optimum.

Using surrogate models in evolutionary search provides an efficient means of handling today's complex applications plagued with increasing high-computational needs. Recent surrogate-assisted evolutionary frameworks have relied on the use of a variety of different modeling approaches to approximate the complex problem landscape. From these recent studies, one main research issue is with the choice of modeling scheme used, which has been found to affect the performance of evolutionary search significantly. Given that theoretical knowledge available for making a decision on an approximation modela prioriis very much limited, this paper describes a generalization of surrogate-assisted evolutionary frameworks for optimization of problems with objectives and constraints that are computationally expensive to evaluate. The generalized evolutionary framework unifies diverse surrogate models synergistically in the evolutionary search. In particular, it focuses on attaining reliable search performance in the surrogate-assisted evolutionary framework by working on two major issues: 1) to mitigate the'curse of uncertainty'robustly, and 2) to benefit from the 'bless of uncertainty.' The backbone of the generalized framework is a surrogate-assisted memetic algorithm that conducts simultaneous local searches usingensembleandsmoothingsurrogate models, with the aims of generating reliable fitness prediction and search improvements simultaneously. Empirical study on commonly used optimization benchmark problems indicates that the generalized framework is capable of attaining reliable, high quality, and efficient performance under a limited computational budget.

The discovery of low-energy stable and meta-stable molecular structures remains an important and unsolved problem in search and optimization. In this paper, we contribute two stochastic algorithms, the archiving molecular memetic algorithm (AMMA) and the archiving basin hopping algorithm (ABHA) for sampling low-energy isomers on the landscapes of pure water clusters (H2O)n. We applied our methods to two sophisticated empirical water cluster models, TTM2.1-F and OSS2, and generated archives of low-energy water isomers (H2O)n n=3-15. Our algorithms not only reproduced previously-found best minima, but also discovered new global minima candidates for sizes 9-15 on OSS2. Further numerical results show that AMMA and ABHA outperformed a baseline stochastic multistart local search algorithm in terms of convergence and isomer archival. Noting a performance differential between TTM2.1-F and OSS2, we analyzed both model landscapes to reveal that the global and local correlation properties of the empirical models differ significantly. In particular, the OSS2 landscape was less correlated and hence, more difficult to explore and optimize. Guided by our landscape analyses, we proposed and demonstrated the effectiveness of a hybrid local search algorithm, which significantly improved the sampling performance of AMMA on the larger OSS2 landscapes. Although applied to pure water clusters in this paper, AMMA and ABHA can be easily modified for subsequent studies in computational chemistry and biology. Moreover, the landscape analyses conducted in this paper can be replicated for other molecular systems to uncover landscape properties and provide insights to both physical chemists and evolutionary algorithmists.

Recent studies have shown that clearing schemes are efficient multi-modal optimization methods. They efficiently reduce genetic drift which is the direct reason for premature convergence in genetic algorithms. However, clearing schemes assumed a landscape containing equal-spaced basins when using a fixed niche radius. Further, most clearing methods employ policies that favor elitists, thus affecting the explorative capabilities of the search. In this paper, we present a valley adaptive clearing scheme, aiming at adapting to non-uniform width of the valleys in the problem landscape. The framework of the algorithm involves hill-valley initialization, valley-adaptive clearing and archiving. Experimental results on benchmark functions are presented to demonstrate that the proposed scheme uncovers more local optima solutions and displays excellent robustness to varying niche radius than other clearing compeers.

A cellular genetic algorithm (CGA) is a decentralized form of GA where individuals in a population are usually arranged in a 2D grid and interactions among individuals are restricted to a set neighborhood. In this paper, we extend the notion of cellularity to memetic algorithms (MA), a configuration termed cellular memetic algorithm (CMA). In addition, we propose adaptive mechanisms that tailor the amount of exploration versus exploitation of local solutions carried out by the CMA. We systematically benchmark this adaptive mechanism and provide evidence that the resulting adaptive CMA outperforms other methods both in the quality of solutions obtained and the number of function evaluations for a range of continuous optimization problems.

Memetic algorithms (MAs) represent one of the recent growing areas in evolutionary algorithm (EA) research. The term MAs is now widely used as a synergy of evolutionary or any population-based approach with separate individual learning or local improvement procedures for problem search. Quite often, MAs are also referred to in the literature as Baldwinian EAs, Lamarckian EAs, cultural algorithms, or genetic local searches. In the last decade, MAs have been demonstrated to converge to high-quality solutions more efficiently than their conventional counterparts on a wide range of real-world problems. Despite the success and surge in interests on MAs, many of the successful MAs reported have been crafted to suit problems in very specific domains. Given the restricted theoretical knowledge available in the field of MAs and the limited progress made on formal MA frameworks, we present a novel probabilistic memetic framework that models MAs as a process involving the decision of embracing the separate actions of evolution or individual learning and analyzing the probability of each process in locating the global optimum. Further, the framework balances evolution and individual learning by governing the learning intensity of each individual according to the theoretical upper bound derived while the search progresses. Theoretical and empirical studies on representative benchmark problems commonly used in the literature are presented to demonstrate the characteristics and efficacies of the probabilistic memetic framework. Further, comparisons to recent state-of-the-art evolutionary algorithms, memetic algorithms, and hybrid evolutionary-local search demonstrate that the proposed framework yields robust and improved search performance.

Evolutionary gradient search is a hybrid algorithm that exploits the complementary features of gradient search and evolutionary algorithm to achieve a level of efficiency and robustness that cannot be attained by either techniques alone. Unlike the conventional coupling of local search operators and evolutionary algorithm, this algorithm follows a trajectory based on the gradient information that is obtain via the evolutionary process. In this paper, we consider how gradient information can be obtained and used in the context of multi-objective optimization problems. The different types of gradient information are used to guide the evolutionary gradient search to solve multi-objective problems. Experimental studies are conducted to analyze and compare the effectiveness of various implementations.

Many deterministic algorithms in the context of constrained optimization require the first-order derivatives, or the gradient vectors, of the objective and constraint functions to determine the next feasible direction along which the search should progress. Although the second-order derivatives, or the Hessian matrices, are also required by some methods such as the sequential quadratic programming (SQP), their values can be approximated based on the first-order information, making the gradients central to the deterministic algorithms for solving constrained optimization problems. In this paper, two ways of obtaining the gradients are compared under the framework of the simple memetic algorithm (MA) employing genetic algorithm (GA) and SQP. Despite the simplicity and straightforwardness of the finite-difference gradients, faster convergence rate can be achieved when the analytical gradients can be made available. The savings on the number of function evaluations as well as the amount of time taken to solve some benchmark problems are presented along with some discussions.

This paper reviews the different gradient-based schemes and the sources of gradient, their availability, precision and computational complexity, and explores the benefits of using gradient information within a memetic framework in the context of continuous parameter optimization, which is labeled here as memetic gradient search. In particular, we considered a quasi-Newton method with analytical gradient and finite differencing, as well as simultaneous perturbation stochastic approximation, used as the local searches. Empirical study on the impact of using gradient information showed that memetic gradient search outperformed the traditional GA and analytical, precise gradient brings considerable benefit to gradient-based local search (LS) schemes. Though gradient-based searches can sometimes get trapped in local optima, memetic gradient searches were still able to converge faster than the conventional GA.

We report our work on the algorithmic development of an evolutionary methodology for automatic configuration of metaheuristic algorithms for solving complex combinatorial optimization problems.We term it Automatic Configuration Engine for Metaheuristics (ACEM). We first propose a novel Left Variation - Right Property (LVRP) tree structure to manage various metaheuristic procedures and properties. With LVRP tree, feasible configurations of metaheuristics can be easily specified. An evolutionary learning algorithm is then proposed to evolve the internal context of the trees based on pre-selected training set. Guided by a user-defined satisfaction function of the candidate algorithms, it converges to the optimal or a very good algorithm. The experimental comparison with two recent state-of-the-art algorithms for solving the quadratic assignment problem (QAP) shows that ACEM produces an hybrid-genetic algorithm with human-competitive or even better performance.

Given the importance of optimization and informatics which are the two broad fields of research, we present an instance of Optinformatics which denotes the specialization of informatics for the processing of data generated in optimization so as to extract possibly implicit and potentially useful information and knowledge. In particular, evolutionary computation does not have to be entirely a black-box approach that generates only the global optimal or good quality solutions. How the solutions are obtained in evolutionary search may be brought to light through Optinformatics. In this paper, we present a Frequent Schemas Analysis (FSA) technique for extracting knowledge from the search process by using the historical optimization data, which are otherwise often discarded. FSA bring about greater understanding of GA dynamics through mining for frequent schemas that exists implicitly within the optimization data via the design of frequent pattern techniques (LoFIA) in informatics. To illustrate the principle of optinformatics, a case study using the Royal Road problem is used to explain the search performance of Genetic Algorithm (GA) in action.

In recent years, there has been an increase in research activities on Memetic Algorithm (MA). MA works with memes; a meme being defined as "the basic unit of cultural transmission, or imitation" [5]. In this respect, a Memetic Algorithm essentially refers to "an algorithm that mimics the mechanisms of cultural evolution". To date, there has been significant effort in bringing MA closer to the idea of cultural evolution. In this paper we assess MAs from the perspectives of "Universal Darwinism" and "Memetics". Subsequently, we propose a Diffusion Memetic Algorithm where the memetic material is transmitted by means of non-genetic transfer. Numerical studies are presented based on some of the commonly used synthetic problems in continuous optimization.

In the paper, the multi-class pattern classification using extreme learning machine (ELM) is studied. The study is based on either a series of ELM binary classifiers or a single ELM classifier. When using binary ELM classifiers, the multi-class problem is decomposed into two-class problem using the one-against-all (OAA) and one-against-one (OAO) schemes, which are named as ELM-OAA and ELM-OAO respectively for brevity. In a single ELM classifier, the multi-class problem is implemented with an architecture of multi-output nodes which is equal to the number of pattern classes. Their performance is evaluated using some multi-class benchmark problems and simulation results show that ELM-OAA and ELM-OAO requires fewer hidden nodes than the single ELM classifier. In addition ELM-OAO usually has similar or less computation burden than the single ELM classifier when the pattern class labels is not larger than 10.

Regression has been successfully combined with the memetic algorithm (MA) for constructing surrogate models. It is essentially an attempt to approximate the objective or constraint landscape of a constrained optimization problem. Classification, on the other hand, has probably never been thought of being of any assistance to the MA. In fact, it can be used to approximate the feasibility boundary by means of some decision functions. The search effort can thus be focussed on the nearby region, recalling that many constrained optimization problems have their optimal solutions situated on the boundaries. This simply means that only potential individuals will undergo local refinements, reducing the number of function evaluations and accelerating the identification of the global optimum. Presented in this paper is a new approach that combines the support vector machine (SVM) with the MA to achieve this purpose.

This paper addresses the collaborative linear- assignment problem (CLAP) for a class of allocation applications. CLAP entails using agents to seek a concurrent allocation of one different object for every agent, to optimize a linear sum efficiency function as their (soft) social goal. Anchoring in the standard framework of automated negotiation allows an original belief-desire-intention (BDI) negotiation model for CLAP to be conceptually separated into a BDI assignment protocol and an adopted strategy. Facilitated by this conceptual separation, the contributions of this paper are as follows: 1) providing a rigorous analysis of the protocol and demonstrating its salient properties and 2) formulating new strategies using a novel idea of cooperative concession. Four different strategies for a negotiation agent and the arbitration agent provide 16 arbitration-negotiation combinations running with the protocol, and these are empirically assessed for their performance profiles in negotiation speed and solution quality. Important findings, including the stability of the protocol in producing better than good enough global allocations and the strategic influence of cooperative concessions on performance, are examined. The significance and practicality of this paper in relation to existing work are also discussed.

The design and engineering of complex materials and products often requires intricate interactions between domain experts in science, material and engineering as well as the utilization of diverse software systems for discovery and optimization. If left as it is, design engineers would most likely be at a loss on how to engage the entire entourage of the multi- disciplinary processes as well as the compute-intensive and data-intensive nature of the activities involved. This paper describes a possible solution through the development of a Grid-based Problem Solving Environment for Engineering of Materials (GPEM). The GPEM aims to provide a one-stop platform where engineers will perform material discovery, design optimization and material characterization, with grid computing as the enabling technology. Upon describing the details of the process workflow and the adopted architecture design, the paper will present the current implementation of GPEM, in the design optimization of fractal structures.

Apart from the need to satisfy several competing objectives, many real-world applications are also sensitive to decision or environmental parameter variation which results in large or unacceptable performance variation. While evolutionary optimization techniques have several advantages over operational research methods for robust optimization, it is rarely studied by the evolutionary multi-objective (MO) optimization community. This paper addresses the issue of robust MO optimization by presenting a robust continuous MO test suite with features of noise-induced solution space, fitness landscape and decision space variation. The work presented in this paper should encourage further studies and the development of more effective algorithms for robust MO optimization.

Over the recent years, there has been increasing research activities made on improving the efficacy of memetic algorithm (MA) for solving complex optimization problems. Particularly, these efforts have revealed the success of MA on a wide range of real world problems. MAs not only converge to high quality solutions, but also search more efficiently than their conventional counterparts. Despite the success and surge in interests on MAs, there is still plenty of scope for furthering our understanding on how and why synergy between population- based and individual learning searchers would lead to successful Memetic Algorithms. In this paper we outline several important design issues of Memetic Algorithms and present a systematic study on each. In particular, we conduct extensive experimental studies on the impact of each individual design issue and their relative impacts on memetic search performances by means of three commonly used synthetic problems. From the empirical studies obtained, we attempt to reveal the behaviors of several MA variants to enhance our understandings on MAs.

Distributed computing environments offer vast amounts of computational power for use in parallel memetic algorithms. However, they consist of heterogeneous computing nodes, in terms of computational power, operating platform, network connectivity and latency. The behavior of parallel memetic algorithms in such environment is poorly understood: the vast majority of current parallel MAs assumes homogeneous environment. To deal with the heterogeneity of the computing resources, a hierarchical model PMA (hPMA-DLS) is proposed to provide the speed-up regardless of the heterogeneity in the distributed environment while preserving the standard behavior of the PMA. The empirical study on several large scale quadratic assignment problems (QAPs) shows that hPMA-DLS can enhance the efficiency of the island model PMA-DLS [22] search without deterioration in the solution quality.

Surrogate-Assisted Memetic Algorithm (SAMA) is a hybrid evolutionary algorithm, particularly a memetic algorithm that employs surrogate models in the optimization search. Since most of the objective function evaluations in SAMA are approximated, the search performance of SAMA is likely to be affected by the characteristics of the models used. In this paper, we study the search performance of using different meta modeling techniques, ensembles, and multi-surrogates in SAMA. In particular, we consider the SAMA-TRF, a SAMA model management framework that incorporates a trust region scheme for interleaving use of exact objective function with computationally cheap local meta models during local searches. Four different metamodels, namely Gaussian Process (GP), Radial Basis Function (RBF), Polynomial Regression (PR), and Extreme Learning Machine (ELM) neural network are used in the study. Empirical results obtained show that while some metamodeling techniques perform best on particular benchmark problems, ensemble of metamodels and multisurrogates yield robust and improved solution quality on the benchmark problems in general, for the same computational budget.

This paper proposes a period representation for modeling the multidrug HIV therapies and an Adaptive Multimeme Algorithm (AMmA) for designing the optimal therapy. The period representation offers benefits in terms of flexibility and reduction in dimensionality compared to the binary representation. The AMmA is a memetic algorithm which employs a list of three local searchers adaptively activated by an evolutionary framework. These local searchers, having different features according to the exploration logic and the pivot rule, have the role of exploring the decision space from different and complementary perspectives and, thus, assisting the standard evolutionary operators in the optimization process. Furthermore, the AMmA makes use of an adaptation which dynamically sets the algorithmic parameters in order to prevent stagnation and premature convergence. The numerical results demonstrate that the application of the proposed algorithm leads to very efficient medication schedules which quickly stimulate a strong immune response to HIV. The earlier termination of the medication schedule leads to lesser unpleasant side effects for the patient due to strong antiretroviral therapy. A numerical comparison shows that the AMmA is more efficient than three popular metaheuristics. Finally, a statistical test based on the calculation of the tolerance interval confirms the superiority of the AMmA compared to the other methods for the problem under study.

The ten papers in this special section are devoted to memetic algorithms. The papers are loosely grouped into two categories: memetic algorithm methodologies and domain-specific memetic algorithms. Briefly summarizes the articles included in this section.

In this paper, we present a novel surrogate-assisted evolutionary optimization framework for solving computationally expensive problems. The proposed framework uses computationally cheap hierarchical surrogate models constructed through online learning to replace the exact computationally expensive objective functions during evolutionary search. At the first level, the framework employs a data-parallel Gaussian process based global surrogate model to filter the evolutionary algorithm (EA) population of promising individuals. Subsequently, these potential individuals undergo a memetic search in the form of Lamarckian learning at the second level. The Lamarckian evolution involves a trust-region enabled gradient-based search strategy that employs radial basis function local surrogate models to accelerate convergence. Numerical results are presented on a series of benchmark test functions and on an aerodynamic shape design problem. The results obtained suggest that the proposed optimization framework converges to good designs on a limited computational budget. Furthermore, it is shown that the new algorithm gives significant savings in computational cost when compared to the traditional evolutionary algorithm and other surrogate assisted optimization frameworks

This correspondence presents a novel hybrid wrapper and filter feature selection algorithm for a classification problem using a memetic framework. It incorporates a filter ranking method in the traditional genetic algorithm to improve classification performance and accelerate the search in identifying the core feature subsets. Particularly, the method adds or deletes a feature from a candidate feature subset based on the univariate feature ranking information. This empirical study on commonly used data sets from the University of California, Irvine repository and microarray data sets shows that the proposed method outperforms existing methods in terms of classification accuracy, number of selected features, and computational efficiency. Furthermore, we investigate several major issues of memetic algorithm (MA) to identify a good balance between local search and genetic search so as to maximize search quality and efficiency in the hybrid filter and wrapper MA.

In both numerical and stochastic optimization methods, surrogate models are often employed in lieu of the expensive high-fidelity models to enhance search efficiency. In gradient-based numerical methods, the trustworthiness of the surrogate models in predicting the fitness improvement is often addressed using ad hoc move limits or a trust region framework (TRF). Inspired by the success of TRF in line search, here we present a Trusted Evolutionary Algorithm (TEA) which is a surrogate-assisted evolutionary algorithm that exhibits the concept of surrogate model trustworthiness in its search. Empirical study on benchmark functions reveals that TEA converges to near-optimum solutions more efficiently than the canonical evolutionary algorithm.

Evolutionary frameworks that employ approximation models or surrogates for solving optimization problems with computationally expensive fitness functions may be referred as Surrogate-Assisted Evolutionary Algorithms (SAEA). In this paper, we present a study on the effects of uncertainty in the surrogate on SAEA. In particular, we focus on both the ‘ curse of uncertainty’ and ‘ blessing of uncertainty’ on evolutionary search, a notion borrowed from ‘ curse and blessing of dimensionality’ in [1]. Here, the ‘ curse of uncertainty’ refers to impairments due to the errors in the approximation. The ‘ blessing of uncertainty’ is less explicitly discussed in the literature, but refers to the benefits of approximation errors on evolutionary search. Empirical studies suggest that approximation errors lead to convergence at false global optima, but prove to be beneficial in some cases.

In this paper, we propose the island model parallel memetic algorithm with diversity-based dynamic adaptive strategy (PMA-DLS) for controlling the local search frequency and demonstrate its utility in solving complex combinatorial optimization problems, in particular large-scale quadratic assignment problems (QAPs). The empirical results show that PMA-DLS converges to competitive solutions at significantly lower computational cost when compared to the canonical MA and PMA. Furthermore, compared to our previous work on PMA using static adaptation strategy, it is found that the diversity-based dynamic adaptation strategy displays better robustness in terms of solution quality across the class of QAP problems considered without requiring extra effort in selecting suitable parameters.

In recent decades, many meta-heuristics, including genetic algorithm (GA), ant colony optimization (ACO) and various local search (LS) procedures have been developed for solving a variety of NP-hard combinatorial optimization problems. Depending on the complexity of the optimization problem, a meta-heuristic method that may have proven to be successful in the past might not work as well. Hence it is becoming a common practice to hybridize meta-heuristics and local heuristics with the aim of improving the overall performance. In this paper, we propose a novel adaptive GA-ACO-LS hybrid algorithm for solving quadratic assignment problem (QAP). Empirical study on a diverse set of QAP benchmark problems shows that the proposed adaptive GA-ACO-LS converges to good solutions efficiently. The results obtained were compared to the recent state-of-the-art algorithm for QAP, and our algorithm showed obvious improvement.

In theoretical computer science, it is well-established fact that Turing model (and its equivalent models like Chomsky's phrase structured grammar model, or recursively enumerable functions) is a most general model for computation. Many biologically inspired computing paradigms such as traditional neural networks, genetic algorithms, evolutionary computing frameworks etc., aim at addressing the problem of automatic generation of algorithmic solutions for many computationally difficult problems. Various soft computing models have complementary capabilities. This motivates us to investigate systematic framework for their integration. In this paper, we give an overview of the computational capabilities of some soft computing models, and make some observations about their integration

Terrain models that permit multiresolution access are essential for model predictive control of unmanned aerial vehicles in low-level flights. The authors present the extreme learning machine (ELM), a recently proposed learning paradigm, as a mechanism for learning the stored digital elevation information to allow multiresolution access. We give results of simulations designed to compare the performance of our approach with two other approaches for multiresolution access, namely: 1) linear interpolation on Delaunay triangles of the sampled terrain data points and 2) terrain learning using support vector machines (SVMs). The results show that to achieve the same mean square error during access, the memory needed in our approach is significantly lower. Additionally, the offline training time for the ELM network is much less than that for the SVM.

Solving design optimization problems using evolutionary algorithms has always been perceived as finding the optimal solution over the entire search space. However, the global optima may not always be the most desirable solution in many real-world engineering design problems. In practice, if the global optimal solution is very sensitive to uncertainties, for example, small changes in design variables or operating conditions, then it may not be appropriate to use this highly sensitive solution. In this paper, we focus on combining evolutionary algorithms with function approximation techniques for robust design. In particular, we investigate the application of robust genetic algorithms to problems with high dimensions. Subsequently, we present a novel evolutionary algorithm based on the combination of a max-min optimization strategy with a Baldwinian trust-region framework employing local surrogate models for reducing the computational cost associated with robust design problems. Empirical results are presented for synthetic test functions and aerodynamic shape design problems to demonstrate that the proposed algorithm converges to robust optimum designs on a limited computational budget.

This paper describes a novel design of a fuzzy inference chip that allows for real-time online context switching. A context refers to a situation or scenario of an application requiring specific domain knowledge. In particular, our focus is on the class of applications involving embedded fuzzy control. The domain knowledge therefore refers to fuzzy rules and memberships. The kind of applications being considered is real-time in nature, which necessitates the implementation of hardware for fuzzy inferencing. The chip architecture is described and details on the design of the chip is presented.

Adaptation of parameters and operators represents one of the recent most important and promising areas of research in evolutionary computations; it is a form of designing self-configuring algorithms that acclimatize to suit the problem in hand. Here, our interests are on a recent breed of hybrid evolutionary algorithms typically known as adaptive memetic algorithms (MAs). One unique feature of adaptive MAs is the choice of local search methods or memes and recent studies have shown that this choice significantly affects the performances of problem searches. In this paper, we present a classification of memes adaptation in adaptive MAs on the basis of the mechanism used and the level of historical knowledge on the memes employed. Then the asymptotic convergence properties of the adaptive MAs considered are analyzed according to the classification. Subsequently, empirical studies on representatives of adaptive MAs for different type-level meme adaptations using continuous benchmark problems indicate that global-level adaptive MAs exhibit better search performances. Finally we conclude with some promising research directions in the area.

Multi-cluster Grids have emerged as the most popular type of grid environments. On multi-cluster grid, resources are distributed across different networks. Hence, scheduling and dispatching jobs are difficult and inefficient due to the restrictions of network and the communication overhead. This paper presents the design and implementation of an efficient GridRPC system for the multi-cluster grid. Metascheduling mechanism is developed to distribute jobs onto different clusters of computers effectively. Further, a dynamic bundling mechanism is provided to reduce overhead of job dispatching. This mechanism bundles similar requests to form a composite request that is dispatched to a grid resource. The above mechanisms have been tested with different types of services and loads, and the results show that they help to achieve more balanced workload and reduce the overhead of job dispatching. This cuts down execution time of applications.

This paper presents a study on hierarchical surrogate-assisted evolutionary algorithm (HSAEA) using different global surrogate models for solving computationally expensive optimization problems. In particular, we consider the use of Gaussian process (GP) and polynomial regression (PR) methods for approximating the global fitness landscape in the surrogate-assisted evolutionary search. The global surrogate model serves to pre-screen the EA population for promising individuals. Subsequently, these potential individuals undergo a local search in the form of Lamarckian learning using online local surrogate models. Numerical results are presented on two multimodal benchmark test functions. The results obtained show that both PR-HSAEA and GP-HSAEA converge to good designs on a limited computational budget. Further, our study also shows that the GP model is suitable for global surrogate modeling in HSAEA if the evaluation function is very expensive in computations. On moderately expensive problems, the PR method may serve to generate better efficiency than using GP.

Memetic algorithms have become to gain increasingly important for solving large scale combinatorial optimization problems. Typically, the extent of the application of local searches in canonical memetic algorithm is based on the principle of "more is better". In the same spirit, the island model parallel memetic algorithm (PMA) is an important extension of the canonical memetic algorithm which applies local searches to every transitional solutions being considered. For PMA which applies complete local search, we termed it as PMA-CLS. In this paper, we consider the island model PMA with selective application of local search (PMA-SLS) and demonstrate its utility in solving complex combinatorial optimization problems, in particular large-scale quadratic assignment problems (QAPs). Based on our empirical results, the PMA-SLS compared to the PMA-CLS, can reduce the computational time spent significantly with little or no lost of solution quality. This we concluded is due mainly to the ability of the PMA-SLS to manage a more desirable diversity profile as the search progresses.

This work presents enhancements to a surrogate-assisted evolutionary optimization framework proposed earlier in the literature for solving computationally expensive design problems on a limited computational budget (Ong et al., 2003). The main idea of our former framework was to couple evolutionary algorithms with a feasible sequential quadratic programming solver in the spirit of Lamarckian learning, including a trust-region approach for interleaving the true fitness function with computationally cheap local surrogate models during gradient-based search. We propose a hierarchical surrogate-assisted evolutionary optimization framework for accelerating the convergence rate of the original surrogate-assisted evolutionary optimization framework. Instead of using the exact high-fidelity fitness function during evolutionary search, a Kriging global surrogate model is used to screen the population for individuals that undergo Lamarckian learning. Numerical results are presented for two multimodal benchmark test functions to show that the proposed approach leads to a further acceleration of the evolutionary search process.

This paper presents the experiences gained from ongoing research collaboration between the School of Computer Engineering at Nanyang Technological University and the Southampton e-Science centre at the University of Southampton using a service-oriented approach for complex engineering design optimisation. The service-oriented approach enables programmatic collaboration to be realized while maintaining the autonomy of individual codes at the different institutes and organizations. In the current work, a genetic algorithm optimisation logic implemented as a grid service at Southampton is used to drive the design search process, while the aerodynamic analysis code located in Singapore is used to evaluate the objective function of the design points. Experience gathered from the current study on airfoil shape optimisation is valuable for establishing effective, efficient and customised programmatic links between institutions to solve complicated engineering design problems.

This paper extends our previous work on the island model parallel hybrid-genetic algorithm (PHGA) for large scale quadratic assignment problems (QAPs). Some issues on the control parameters of the migration process and how they affect the quality of the solutions and the efficiency of algorithm deserve further evaluative study. In this paper, we investigate the effect of migration topology on the performance of the PHGA. Two topologies, one-way ring topology and random topology, are studied and analyzed. The empirical results show that the PHGA with ring topology is better able to achieve an appropriate tradeoff between exploration and exploitation and hence more helpful to improve the performance of PHGA for solving large scale QAPs.

Over the last decade, memetic algorithms (MAs) have relied on the use of a variety of different methods as the local improvement procedure. Some recent studies on the choice of local search method employed have shown that this choice significantly affects the efficiency of problem searches. Given the restricted theoretical knowledge available in this area and the limited progress made on mitigating the effects of incorrect local search method choice, we present strategies for MA control that decide, at runtime, which local method is chosen to locally improve the next chromosome. The use of multiple local methods during a MA search in the spirit of Lamarckian learning is here termed Meta-Lamarckian learning. Two adaptive strategies for Meta-Lamarckian learning are proposed in this paper. Experimental studies with Meta-Lamarckian learning strategies on continuous parametric benchmark problems are also presented. Further, the best strategy proposed is applied to a real-world aerodynamic wing design problem and encouraging results are obtained. It is shown that the proposed approaches aid designers working on complex engineering problems by reducing the probability of employing inappropriate local search methods in a MA, while at the same time, yielding robust and improved design search performance.

In this paper, we consider a parallel hybrid-GA (PHGA) for solving large combinatorial optimization problem. The approach of our PHGA is based on the island model whereby islands of subpopulations are farmed to multiple processing nodes for execution. The PHGA was applied to solve the quadratic assignment problems (QAP) to demonstrate the potential effectiveness of the models. In particular, we concentrate on QAP benchmarks of high complexity for n ranging from 60 to 256. Our results show that a two-island PHGA which employs a simplistic elite migration between islands outperforms the serial hybrid-GA (SHGA) significantly. As the size and complexity of the problem increases, the advantage of the PHGA in terms of computation time and solution quality becomes more evident. This opens up a wide channel for further exploration on implementation of the PHGA in a grid computing environment.

In this paper, we present an evolutionary framework for efficient aerodynamic shape design. The approach suggests employing hybrid evolutionary algorithm with gradient-based local search method in the spirit of Lamarckian and surrogate models that approximates the computationally expensive adjoint computational fluid dynamics during design search. In particular, we reveal that the proposed framework guarantees global convergence by inheriting the properties of trust-region method to interleave use of the exact solver for the objective function with computationally cheap surrogate models during local search. Empirical results on 2D airfoil shape design using an adjoint inverse pressure design problem indicates that the approaches global convergences on a limited computational budget.

This paper describes a kernel active handwriting model (K-AHM) and its application to handwritten Chinese character recognition. In the model, the kernel principal component analysis is applied to capture nonlinear variations caused by handwriting, and a fitness function on the basis of a chamfer distance transform is introduced to search for optimal shape parameters using genetic algorithms (GAs). The K-AHM is applied to handwritten Chinese character recognition, which converts the complex pattern recognition problem into recognizing a small set of primitive structures called radicals. By treating Chinese character composition as a discrete-time Markov process, character composition is carried out with the Viterbi algorithm. The proposed methodology has been successfully implemented in an experimental recognition system.