Large language models (LLMs) show promising performance on small-scale graph reasoning tasks but fail when handling real-world graphs with complex queries. This phenomenon arises from LLMs' working memory constraints, which result in their inability to retain long-range graph topology over extended contexts while sustaining coherent multi-step reasoning. However, real-world graphs are often structurally complex, such as Web, Transportation, Social, and Citation networks. To address these limitations, we propose GraphCogent, a collaborative agent framework inspired by human Working Memory Model that decomposes graph reasoning into specialized cognitive processes: sense, buffer, and execute. The framework consists of three modules: Sensory Module standardizes diverse graph text representations via subgraph sampling, Buffer Module integrates and indexes graph data across multiple formats, and Execution Module combines tool calling and tool creation for efficient reasoning. We also introduce Graph4real, a comprehensive benchmark that contains four domains of real-world graphs (Web, Transportation, Social, and Citation) to evaluate LLMs' graph reasoning capabilities. Our Graph4real covers 21 different graph reasoning tasks, categorized into three types (Structural Querying, Algorithmic Reasoning, and Predictive Modeling tasks), with graph scales up to 10 times larger than existing benchmarks. Experiments show that Llama3.1-8B based GraphCogent achieves a 50% improvement over massive-scale LLMs like DeepSeek-R1 (671B). Compared to state-of-the-art code-based baseline, our framework outperforms by 20% in accuracy while reducing token usage by 80% for in-toolset tasks and 30% for out-toolset tasks.

Recommendation systems alleviate the issue of information overload via modeling user preferences from interaction sequences. Although self-attention based sequential models effectively capture long-range dependencies, they are susceptible to noise amplification in sparse sequences and over-smoothing of item representations, which obscures true user intent and reduces sensitivity to fine-grained behavioral changes. To overcome these challenges, we propose SAFA, a sparse sequential recommendation framework comprising: (1) an adaptive sparse attention mechanism that suppresses noisy interactions while preserving embedding diversity; (2) a frequency-aware encoder that decomposes interaction sequences into low-frequency components for long-term preference modeling and high-frequency components for short-term intent dynamics; and (3) a simplified focal loss that removes the class-balancing term while preserving the focusing factor, emphasizing hard-to-predict samples rather than class priors. Experiments on seven benchmark datasets demonstrate that SAFA consistently achieve state-of-the-art performance with average improvements of up to 3.77%, 4.10% and 4.25% in terms of HR@5, HR@10 and HR@20, respectively, and 4.30%, 4.78% and 4.58% in terms of NDCG@5, NDCG@10 and NDCG@20, respectively, over the best competing model. Ablation studies verify the importance of each component, with notable performance degradation upon removing the sparse attention or frequency-aware encoder. Overall, SAFA enhances sequential recommendation by improving robustness and discriminative learning under noisy and sparse conditions.

Multi-layer network community detection is a crucial topic in Industrial Internet of things(IIoT). Due to communication and privacy requirements, network data is distributed across multiple devices, being a significant challenge to develop a model to learn latent information for community detection. To address the problem, this paper proposes a federated fuzzy C-Means for multi-layer network community detection. Firstly, non-negative matrix factorization is employed to obtain a low-dimensional representation via training local data in each client. The gradients of the global centroids are then transmitted to a central server for consistent fusion and complete community detection within the fuzzy C-Means framework. As a result, the training process for each client remains independent and leads to effectively privacy preservation. Experimental results demonstrate that the proposed method can successfully perform multi-layer network community detection across distributed devices and achieve comparable performance in contrast with centralized community detection methods on four public datasets.

Cancer subtyping is crucial for categorizing patients into distinct groups, enabling precision medicine and personalized therapies. As multi-omic analysis becomes more prevalent, integrating data from various omics provides deeper insights into the potential relationships between cancer subtypes. Although most cancer subtyping methods show promising performance, they have several limitations. These methods fail to account for omic differences, address noise in similarity matrices, and preserve the manifold structure of high-dimensional data in lowdimensional space. This study proposes a Robust Diverse Multiview Learning (RDML) model for cancer subtyping. Specifically, multi-view self-representation matrices are formulated as a thirdorder tensor. Differences between views are captured using an orthogonal diversity term, thereby reducing the redundant information between views. To enhance robustness of model to noise, we explicitly separate the self-representation tensor into a clean tensor and a noise tensor. Additionally, Laplacian manifold regularization is employed to preserve the local structure of highdimensional data in low-dimensional space. An efficient algorithm is designed to solve the proposed model. Comprehensive experiments are conducted on ten datasets, demonstrating the superior performance of the proposed model.

The rapid growth of digital interaction has produced a massive volume of complex, high-dimensional data. This data reflects individual behavior, group dynamics, and large-scale patterns across social platforms, sensor networks, and digital infrastructures. Traditional methods struggle to keep up with the speed and scale of this data. New models are needed to find structure, uncover hidden patterns, and support timely decisions [1]. As digital and physical environments continue to overlap, through smart cities, wearables, and connected devices, computational social systems must adapt to handle behaviors that are increasingly fast-changing, distributed, and varied across data types.

Unbalanced incomplete multiview data are widely generated in engineering areas due to sensor failures, data acquisition limitations, etc. However, current research works are rarely focused on unbalanced incomplete multiview unsupervised feature selection (MUFS). To address this issue, this article proposes an MUFS method called unbalanced incomplete multiview unsupervised feature selection with low-redundancy constraint in low-dimensional space (UIMUFSLR). Specifically, the proposed method mitigates the impact of missing samples by learning a unified graph with assigning weights of samples adaptively. In addition, a novel regularization is designed by utilizing the inner product of selected features to obtain low redundancy. An iterative optimization algorithm is devised for UIMUFSLR, accompanied by a comprehensive analysis of its convergence behavior and computational complexity. Experimental results demonstrate the competitiveness of UIMUFSLR in handling unbalanced incomplete multiview data on seven public datasets.

Graph-based multiview clustering (MVC) approaches have demonstrated impressive performance by leveraging the consistency properties of multiview data in an unsupervised manner. However, existing methods for graph learning heavily rely on either Euclidean structures or the manifold topological structures derived from fixed view-specific graphs. Unfortunately, these approaches may not accurately reflect the consensus topological structure in a multiview setting. To address this limitation and enhance the intrinsic graph learning process, an adaptive exploration of a more appropriate consistency topological structure is required. Toward this end, we propose a novel approach called collaborative topological graph learning (CTGL) for MVC. The key idea is to adaptively discover the consistent topological structure to guide intrinsic graph learning. We achieve this by introducing an auxiliary consistency graph that formulates the topological relevance learning function. However, estimating the auxiliary consistency graph is not straightforward, as it is based on the learned view-specific graphs and requires prior availability. To overcome this challenge, we develop a collaborative learning strategy that simultaneously learns both the auxiliary consistency graph and view-specific graphs using tensor learning techniques. This strategy enables the adaptive exploration of the consistency topological structure during graph learning, resulting in more accurate clustering outcomes. Extensive experiments are provided to show the effectiveness of the proposed method. The source code can be found at https://github.com/CLiu272/CTGL.

Social media has a significant influence, which greatly facilitates people to stay up-to-date with information. Unfortunately, a great deal of fake news on social media misleads people and causes a lot of losses. Therefore, fake news detection is necessary to address this issue. Recently, social content category-based methods have become a crucial component of fake news detection. Different from news context-based category, which focuses on word embedding, it tends to explore the potential relationships and structures between users and news. In this article, a third-order tensor, which obtains massive information and connections, is constructed by the social links and engagements of social networks. Then, a sparse and graph-regularized CANDECOMP/PARAFAC (SGCP) tensor decomposition learning method is proposed for fake news detection on social network. In SGCP, a news factor matrix is constructed by CP decomposition of the tensor, which reflects the complex connections among users and news. Furthermore, SGCP retains sparsity of the news factor matrix and preserves the manifold structures from the original space. In addition, an efficient optimization algorithm, which is proven to be monotonically nonincreasing, is proposed to solve SGCP. Finally, abundant experiments are conducted on real-world datasets and demonstrate the effectiveness of the proposed SGCP.

Multiview clustering performs grouping a set of objects by utilizing complementary information from multiple views. Unfortunately, the clustering performance degenerates dramatically if the views are corrupted by noise. To overcome this limitation, we propose a robust multiview clustering approach based on weighted low-rank tensor approximation and noise separation. The proposed model improves the performance through a low-rank approximation function and weighted singular values. The weighted low-rank tensor approximation method considers both prior knowledge and the physical meanings associated with different singular values, leading to superior performance in capturing high-order correlations. Additionally, to eliminate mixed noise, a novellCauchy,1norm is developed to handle outliers, and thel1and Frobenius norms are used to handle random corruptions and slight perturbations, respectively. A high-efficiency optimization algorithm based on the alternating direction method of multipliers (ADMM) is designed to address the challenging proposed model. Experimental results on nine real-world datasets show that the proposed approach outperforms eight state-of-the-art multiview methods. Furthermore, experiments on various kinds of noise demonstrate the superior robustness of the proposed approach. Especially, in the mixed noise condition, the proposed approach is significantly superior to other methods.

With the booming adoption of Electric Vehicles (EVs) globally, the need for reliable and resilient EV Charging Monitoring (EVCM) systems has become crucial. A major challenge in real-time EVCM is the handling of missing data caused by unexpected events, which can impair both real-time monitoring and its downstream applications. To address this vital yet underexplored issue, we propose a curriculum guided multi-feature fusion transformer (CurriFusFormer) learning framework – a novel approach designed to enhance the resilience of EVCM systems against real-time information omissions. Our framework integrates curriculum learning with a multi-feature fusion transformer model, capable of handling various patterns and rates of missing data, ranging from random to block omissions. This innovative approach leverages spatial, temporal, and static features to generate accurate real-time estimations for missing values in diverse scenarios. Extensive experiments on a real-world EVCM dataset demonstrate that CurriFusFormer can perform well with$R^{2}$ranging from 0.92 to 0.83 given the rising missing rate from 30-90%, outperforming seven popular and state-of-the-art methods, especially in scenarios with high missing rates and complex patterns, such as, at 90% missing rate, kNN ($R^{2} =0.65$), XGBoost ($R^{2} =0.78$), BRITS ($R^{2} =0.79$), TFT ($R^{2} =0.80$), and GRIN ($R^{2} =0.82$). All results suggest that the proposed framework could be a promising solution for developing future resilient EVCM networks.

Incomplete multi-view clustering (IMVC) presents a significant challenge due to the need for effectively exploring complementary and consistent information within the context of missing views. One promising strategy to tackle this challenge is to recover missing views by inferring the missing samples. However, such approaches often fail to fully utilize discriminative structural information or adequately address consistency, as it requires such information to be known or learnable in advance, which contradicts the incomplete data setting. In this study, we propose a novel approach calledLatentStructure-Aware view recovery (LaSA) for the IMVC task. Our objective is to recover missing views through discriminative latent representations by leveraging structural information. Specifically, our method offers a unified closed-form formulation that simultaneously performs missing data inference and latent representation learning, using a learned intrinsic graph as structural information. This formulation, incorporating graph structure information, enhances the inference of missing data while facilitating discriminative feature learning. Even when intrinsic graph is initially unknown due to incomplete data, our formulation allows for effective view recovery and intrinsic graph learning through an iterative optimization process. To further enhance performance, we introduce an iterative consistency diffusion process, which effectively leverages the consistency and complementary information across multiple views. Extensive experiments demonstrate the effectiveness of the proposed method compared to state-of-the-art approaches.

This article addresses decentralized robust portfolio optimization based on multiagent systems. Decentralized robust portfolio optimization is first formulated as two distributed minimax optimization problems in a Markowitz return-risk framework. Cooperative-competitive multiagent systems are developed and applied for solving the formulated problems. The multiagent systems are shown to be able to reach consensuses in the expected stock prices and convergence in investment allocations through both intergroup and intragroup interactions. Experimental results of the multiagent systems with stock data from four major markets are elaborated to substantiate the efficacy of multiagent systems for decentralized robust portfolio optimization.

Public buildings such as shopping arcades and railway stations are environments in which pedestrian movement is of significance to many smart building applications. The data-driven approach of pedestrian trajectory prediction is effective in learning a reliable model that can represent complex human movement. Pedestrian trajectories are highly linked to the locations of facilities and services inside a building as pedestrians move towards these destinations for engagement. This paper suggests that the notion of destination is a strong predictor of pedestrian trajectories and proposes a novel enhancement of the data-driven approach for pedestrian tracking in public buildings. The method of destination-driven pedestrian trajectory prediction (DDPTP) first evaluates the most likely destinations of the pedestrian using the destination classifier (DC) and then predicts the future trajectories with the destination-specific trajectory model (DTM). The proposed solution has been evaluated on the NYGC and the ATC datasets and found to outperform state-of-the-art models. The notion of destination can be further developed into a region of interest of which the within-region and out-of-region features can be factored out for more effective learning.

In this paper, a modified Competitive Mechanism Multi-Objective Particle Swarm Optimization (MCMOPSO) algorithm is presented for multi-objective optimization. The algorithm consists of an improved leader selection scheme called multi-competition leader selection. Under this scheme, particles move to the winner among the elite particles for the social cognitive by comparing the nearest angle or the farthest angle of several randomly selected elite particles. Besides, as the inertia weight plays an important role in controlling the previous velocity of each particle, the competitive mechanism is applied to the inertia weight in order to investigate for the most suitable balance between the exploration and exploitation abilities of the algorithm during the search process. The experimental results show that the proposed algorithm outperforms four other popular multi-objective particle swarm optimization algorithms most of the time on thirty-seven benchmarks in terms of inverted generational distance. Furthermore, the proposed algorithm is applied to the signalized traffic problem to optimize the effective green time of each phase, and the proposed algorithm performs better than other MOPSO algorithms for the traffic problem in terms of hypervolume.

Portfolio selection is one of the important issues in financial investments. This article is concerned with portfolio selection based on collaborative neurodynamic optimization. The classic Markowitz mean-variance (MV) framework and its variant mean conditional value-at-risk (CVaR) are formulated as minimax and biobjective portfolio selection problems. Neurodynamic approaches are then applied for solving these optimization problems. For each of the problems, multiple neural networks work collaboratively to characterize the efficient frontier by means of particle swarm optimization (PSO)-based weight optimization. Experimental results with stock data from four major markets show the performance and characteristics of the collaborative neurodynamic approaches to the portfolio optimization problems.

A hybrid algorithm is proposed for multiobjective optimization in this paper. The proposed algorithm consists of multiobjective evolutionary algorithm based on decomposition (MOEA/D) and recurrent neural network, where MOEA/D is for global search while recurrent neural network is for local search. The performance of the proposed algorithm is compared with other three multi-objective algorithms in terms of hypervolume and inverted generational distance. The performance investigation shows that the proposed algorithm generally outperforms the compared algorithms.

There are two ultimate goals in multiobjective optimization. The primary goal is to obtain a set of Pareto-optimal solutions while the secondary goal is to obtain evenly distributed solutions to characterize the efficient frontier. In this paper, a collaborative neurodynamic approach to multiobjective optimization is presented to attain both goals of Pareto optimality and solution diversity. The multiple objectives are first scalarized using a weighted Chebyshev function. Multiple projection neural networks are employed to search for Pareto-optimal solutions with the help of a particle swarm optimization (PSO) algorithm in reintialization. To diversify the Pareto-optimal solutions, a holistic approach is proposed by maximizing the hypervolume (HV) using again a PSO algorithm. The experimental results show that the proposed approach outperforms three other state-of-the-art multiobjective algorithms (i.e., HMOEA/D, MOEA/DD, and NSGAIII) most of times on 37 benchmark datasets in terms of HV and inverted generational distance.

Partitional Clustering is one of the major techniques in Unsupervised Learning in which similar data are put into the same partition. Besides partitioning the unlabeled data, determining the optimal number of partitions is also another main concern in the field of data clustering. Automatic Clustering Differential Evolution (ACDE) is one of the state-of-the-art algorithms that address this concern. In ACDE, the mechanism to determine the optimal number of clusters is by encoding the activation value of each cluster centroid into the chromosome with fixed threshold value. However, it could be argued that a fixed threshold value would be seen as arbitrary, but a varying and adaptive threshold value could yield a solution that would better reflect the quality of clusters. In this paper, a new changing schema of threshold values is introduced for adaptively activating the clusters in the chromosomes, and a heuristic approach is implemented for adjusting the threshold values of each cluster according to their individual quality measurements. The results of several experiments show that the proposed algorithm performed generally better than other state-of-the-art automatic evolutionary clustering algorithms.

Multi-Objective Evolutionary Algorithm based on decomposition (MOEA/D) has been proposed for a decade in solving complex multi-objective problems (MOP) by decomposing it into a set of single-objective problems. MOEA/D-AGR is one of the improved algorithm introduced recently to substitute the original replacement scheme in MOEA/D with a new adaptive global replacement (GR) scheme so that the neighborhood replacement size Tris increased among the generation to achieve the shifting of focus from solution diversity to convergence. However, the new replacement scheme only considers the one-way convergence of the objective solutions among all sub-problems. It is hard to re-achieve the solution diversity once the algorithm reaches another steeper landscape of solution from a flatten one while it is focusing on convergence. This paper proposes a new adaptive GR scheme to prolong the period of Trincrement so that it can fit the re-increment of fitness landscape. To compensate the shorten period for solution convergence, local searching is adapted for those individuals which has stopped improving its solution value by Simulated Annealing (SA) algorithm. In order to suppress the degree of local searching at the early stage of focusing solution diversity, Fuzzy Logic is used here to coordinate the frequency of local searching according to the average change of objective solution values. To demonstrate the performance of the proposed algorithm, several common benchmark MOPs are used in this paper for comparing with the several state-of-the-art MOEA/D algorithms in terms of IGD. The performance investigation found that the performance of the proposed algorithm was generally better than the other compared algorithms.

This paper presents a new Multi-Objective Particle Swarm Optimization (MOPSO) algorithm that has two new components: leader selection and crossover. The new leader selection algorithm, called Space Expanding Strategy (SES), guides particles moving to the boundaries of the objective space in each generation so that the objective space can be expanded rapidly. Besides, crossover is adopted instead of mutation to enhance the convergence and maintain the stability of the generated solutions (exploitation). The performance of the proposed MOPSO algorithm was compared with three popular multi-objective algorithms in solving fifteen standard test functions. Their performance measures were hypervolume, spread and inverse generational distance. The performance investigation found that the performance of the proposed algorithm was generally better than the other three, and the performance of the proposed crossover was generally better than three popular mutation operators.

This paper presents a new algorithm that extends Particle Swarm Optimization (PSO) to deal with multi-objective problems. It makes two main contributions. The first is that the square root distance (SRD) computation among particles and leaders is proposed to be the criterion of the local best selection. This new criterion can make all swarms explore the whole Pareto-front more uniformly. The second contribution is the procedure to update the archive members. When the external archive is full and a new member is to be added, an existing archive member with the smallest SRD value among its neighbors will be deleted. With this arrangement, the non-dominated solutions can be well distributed. Through the performance investigation, our proposed algorithm performed better than two well-known multi-objective PSO algorithms, MOPSO-σ and MOPSO-CD, in terms of different standard measures.