Radiology Report Generation (RRG) aims to automatically generate diagnostic reports from radiology images. To achieve this, existing methods have leveraged the powerful cross-modal generation capabilities of Multimodal Large Language Models (MLLMs), primarily focusing on optimizing cross-modal alignment between radiographs and reports through Supervised Fine-Tuning (SFT). However, by only performing instance-level alignment with the image-text pairs, the standard SFT paradigm fails to establish anatomically-grounded alignment, where the templated nature of reports often leads to sub-optimal generation quality. To address this, we propose S2D-Align, a novel SFT paradigm that establishes anatomically-grounded alignment by leveraging auxiliary signals of varying granularities. S2D-Align implements a shallow-to-deep strategy, progressively enriching the alignment process: it begins with the coarse radiograph-report pairing, then introduces reference reports for instance-level guidance, and ultimately utilizes key phrases to ground the generation in specific anatomical details. To bridge the different alignment stages, we introduce a memory-based adapter that empowers feature sharing, thereby integrating coarse and fine-grained guidance. For evaluation, we conduct experiments on the public MIMIC-CXR and IU X-Ray benchmarks, where S2D-Align achieves state-of-the-art performance compared to existing methods. Ablation studies validate the effectiveness of our multi-stage, auxiliary-guided approach, highlighting a promising direction for enhancing grounding capabilities in complex, multi-modal generation tasks.

Large language models (LLMs) exhibit logically inconsistent hallucinations that appear coherent yet violate reasoning principles, with recent research suggesting an inverse relationship between causal reasoning capabilities and such hallucinations. However, existing reasoning approaches in LLMs, such as Chain-of-Thought (CoT) and its graph-based variants, operate at the linguistic token level rather than modeling the underlying causal relationships between variables, lacking the ability to represent conditional independencies or satisfy causal identification assumptions. To bridge this gap, we introduce causal-DAG construction and reasoning (CDCR-SFT), a supervised fine-tuning framework that trains LLMs to explicitly construct variable-level directed acyclic graph (DAG) and then perform reasoning over it. Moreover, we present a dataset comprising 25,368 samples (CausalDR), where each sample includes an input question, explicit causal DAG, graph-based reasoning trace, and validated answer. Experiments on four LLMs across eight tasks show that CDCR-SFT improves the causal reasoning capability with the state-of-the-art 95.33% accuracy on CLADDER (surpassing human performance of 94.8% for the first time) and reduces the hallucination on HaluEval with 10% improvements. It demonstrates that explicit causal structure modeling in LLMs can effectively mitigate logical inconsistencies in LLM outputs.

Modern applications provided personalized recommendations across diverse scenarios, including the homepage, local pages, and live streams on platforms like TikTok. These scenarios exhibit varying user behavior patterns, resulting in heterogeneous yet interrelated distributions. Existing Multi-Scenario Recommendation (MSR) methods usually use parameter-sharing networks for shared features and scenario-specific networks for unique features. However, these methods fail to handle different distribution across scenarios, resulting in representation entanglement and localization, which hinder effective knowledge transfer and compromise performance. In this paper, we propose a Distribution-aware Re-representations (DAR) method for MSR. Its core idea is to construct distribution-aware prototype spaces and learn disentangled re-representations around global prototypes. Specifically, DAR employs a Multi-gate Mixture of Experts (MMoE) to obtain scenario-shared representations, and uses independent networks to learn scenario-specific representations. These representations are then projected into scenario-shared and scenario-specific prototype spaces, producing scenario-shared re-representations (capturing global information) and scenario-specific re-representations (focusing on distributional differences). During this process, DAR utilizes Unbalanced Optimal Transport (UOT) to compute the transport relationships between representations and global prototypes, taking these as pseudo-labels for re-representation learning. Moreover, to prevent prototype entanglement, a matrix orthogonalization constraint ensures independence among global prototypes. The effectiveness of DAR is demonstrated through extensive offline experiments conducted on four datasets, as well as online A/B tests on a video platform.

Meme-based social abuse detection is challenging because harmful intent often relies on implicit cultural symbolism and subtle cross-modal incongruence. Prior approaches, from fusion-based methods to in-context learning with Large Vision-Language Models (LVLMs), have made progress but remain limited by three factors: i) cultural blindness (missing symbolic context), ii) boundary ambiguity (satire vs. abuse confusion), and iii) lack of interpretability (opaque model reasoning). We introduce CROSS-ALIGN+, a three-stage framework that systematically addresses these limitations: (1) Stage I mitigates cultural blindness by enriching multimodal representations with structured knowledge from ConceptNet, Wikidata, and Hatebase; (2) Stage II reduces boundary ambiguity through parameter-efficient LoRA adapters that sharpen decision boundaries; and (3) Stage III enhances interpretability by generating cascaded explanations. Extensive experiments on five benchmarks and eight LVLMs demonstrate that CROSS-ALIGN+ consistently outperforms state-of-the-art methods, achieving up to 17% relative F1 improvement while providing interpretable justifications for each decision.

Federated learning (FL) enables collaborative model training across distributed data silos, but non-independent and identically distributed (non-IID) heterogeneity among clients often leads to biased or underperforming global models. Neural Architecture Search (NAS) can adapt model structures to diverse data distributions, yet conventional two-stage search-retrain pipelines remain prohibitively expensive for federated settings. We propose OSMONAS (One-Stage Meta-Optimization NAS), a unified framework that integrates Model-Agnostic Meta-Learning (MAML) with differentiable NAS and a progressive soft pruning strategy. OSMONAS jointly optimizes model weights and architectures within a single-stage search, accelerating convergence while preserving performance stability. The integrated pruning mechanism removes redundant operations during search, eliminating the need for post-hoc retraining. Evaluations on standard benchmarks under varying non-IID conditions demonstrate that OSMONAS consistently achieves higher accuracy and faster convergence than state-of-the-art baselines. Beyond computational gains, OSMONAS offers a pathway toward equitable and inclusive federated healthcare, where adaptable architectures enable collaboration across institutions with diverse data, devices, and resources.

Accurately estimating subsurface tumor depth during fluorescence-guided surgery remains an open challenge, as current intraoperative methods provide only surface-level fluorescence contrast without quantitative depth information. Spatial Frequency Domain Imaging offers a pathway toward quantitative fluorescence imaging by capturing both reflectance and optical property maps; however, its application to tumor depth estimation is limited by the scarcity of patient-derived datasets and significant domain gaps between simulated and experimental measurements. To address these challenges, we propose U-KAN, a hybrid deep learning framework that combines a siamese attention U-Net for spatial feature extraction with a Kolmogorov-Arnold Network (KAN) regression head for functional depth mapping. The U-Net captures morphological and structural cues from fluorescence and optical property inputs, while the KAN performs nonlinear pixelwise regression to improve generalization under limited data. Experiments on diffusion-theory, Monte Carlo, and patient-derived phantom datasets demonstrate that the hybrid model achieves more accurate and robust tumor depth estimation than either model alone, establishing a promising foundation for quantitative, depthaware fluorescence imaging in surgical oncology. U-KAN reduced tumor-region MAE by more than 30 % and cut minimum-depth errors nearly in half on phantom data, demonstrating markedly improved cross-domain robustness.

Video anomaly detection (VAD) is crucial in scenarios such as surveillance and autonomous driving, where timely detection of unexpected activities is essential. Albeit existing methods have primarily focused on detecting anomalous objects in videos—either by identifying anomalous frames or objects—they often neglect finer-grained analysis, such as anomalous pixels, which limits their ability to capture a broader range of anomalies. To address this challenge, we propose an innovative VAD framework called Track Any Anomalous Object (TAO), which introduces a Granular Video Anomaly Detection Framework that, for the first time, integrates the detection of multiple fine-grained anomalous objects into a unified framework. Unlike methods that assign anomaly scores to every pixel at each moment, our approach transforms the problem into pixel-level tracking of anomalous objects. By linking anomaly scores to subsequent tasks such as image segmentation and video tracking, our method eliminates the need for threshold selection and achieves more precise anomaly localization, even in long and challenging video sequences. Experiments on extensive datasets demonstrate that TAO achieves state-of-the-art performance, setting a new progress for VAD by providing a practical, granular, and holistic solution. For more information, visit the project page at: https://tao-25.github.io/

Precise mapping of crop types and estimating yields are important in gauging agricultural diversity and yield potential, especially in regions dominated by small-scale farming. Nevertheless, these tasks are challenging due to factors such as small field sizes, inter-cropping, and a lack of sufficient ground truth labels for certain regions. In this paper, we propose an approach that combines advanced deep learning algorithms with Sentinel-2 and MODIS satellite data for improving the accuracy of crop type mapping and yield prediction. We used datasets from the main growing season of 2017 in Kenya (Bungoma, Busia and Siaya) coupled with county level yield data from US, Argentina and Brazil spanning from 2005 to 2016. Our models (CNN, SegNet, MaskRCNN, ResNet, UNet) were evaluated on both tasks i.e., classification of crop types and predicting yields.

Audio deepfakes represent a risk to society as they can deteriorate society’s trust in any audio. In this paper, we present a novel approach for audio deepfake detection using Generative Adversarial Networks (GANs) and contrastive learning in a multi-stage detection framework. In our process, we apply the Pre-trained Models (PTM) to extract all suitable audio phonetics, speaker identity, and other spatial prosodic features or contents, which are crucial for the model. We enhance the model’s performance by utilizing a GAN data augmentation strategy in combination with HiFi-GAN. The Contrastive learning approach is then used for improving the model’s ability to discriminate real speech from fake speech. Our experiments demonstrate that this method is superior to existing methodologies in detection and robustness.

Accurately predicting user watch-time is crucial for enhancing user stickiness and retention in video recommendation systems. Existing watch-time prediction approaches typically involve transformations of watch-time labels for prediction and subsequent reversal, ignoring both the natural distribution properties of label and the instance representation confusion that results in inaccurate predictions. In this paper, we propose ProWTP, a two-stage method combining prototype learning and optimal transport for watch-time regression prediction, suitable for any deep recommendation model. Specifically, we observe that the watch-ratio (the ratio of watch-time to video duration) within the same duration bucket exhibits a multimodal distribution. To facilitate incorporation into models, we use a hierarchical vector quantised variational autoencoder (HVQ-VAE) to convert the continuous label distribution into a high-dimensional discrete distribution, serving as credible prototypes for calibrations. Based on this, ProWTP views the alignment between prototypes and instance representations as a Semi-relaxed Unbalanced Optimal Transport (SUOT) problem, where the marginal constraints of prototypes are relaxed. And the corresponding optimization problem is reformulated as a weighted Lasso problem for solution. Moreover, ProWTP introduces the assignment and compactness losses to encourage instances to cluster closely around their respective prototypes, thereby enhancing the prototype-level distinguishability. Finally, we conducted extensive offline experiments on two industrial datasets, demonstrating our consistent superiority in real-world application.

As latency-sensitive IoT applications proliferate, edge computing becomes critical to sustaining real-time performance. Yet, limited edge infrastructure and reliance on costly static task profiling constrain its potential. This paper presents fReeLoaders, an IoT ecosystem addressing both challenges through opportunistic offloading and adaptive scheduling. fReeLoaders uses nearby idle smart devices to augment edge availability and uses a deadline-driven reinforcement learning scheduler to learn task behavior on the fly, eliminating expensive a priori profiling. Evaluation on real hardware shows it improves quality of service by 11.4% over a state-of-the-art profiling scheduler and adapts to dynamic workloads.

Large-scale recommendation systems mainly rely on sparse ID features, struggling with data sparsity. It's important to use multimodal information to assist ID learning for better performance. However, there exists two challenges: (1) distribution discrepancy between multimodal and ID makes direct integration prone to user-item mismatch; (2) slower convergence of multimodal representations compared to ID, causing optimization imbalance under a unified objective, which limits the potential of multimodal representations. In this paper, we comprehensively investigate the two problems and proposes a framework named AB-Rec to align and balance ID and multimodal representations learning for recommendation. We design three alignment tasks to fine-tune a pre-trained multimodal large language model (MLLM), which is then utilized to generate a unified multimodal representation for each item. AB-Rec aligns the distributions of ID and multimodal representations by minimizing the in-batch Wasserstein distance, and maximizes the distance between the two types of representations for the same item to avoid representation collapse. To solve the optimization imbalance, we propose a gradient modulation method that adaptively controls the optimization process by monitoring the contribution differences between ID and multimodal representations. Finally, we conduct extensive offline experiments on four datasets and an A/B test on an online video platform, demonstrating the effectiveness and scalability of our proposed method.

In video recommendation systems, accurately predicting watch time is crucial for enhancing user engagement and retention. Traditional methods typically apply label transformations or mitigate duration bias to improve performance but overlook that erroneous instance representations are the primary cause of significant prediction errors. Moreover, these approaches predominantly rely on point perdition, limiting their robustness. To address these challenges, we propose LDA, a novel prediction paradigm that optimizes instance representations by explicitly leveraging label distributions as anchors within the model, enabling more accurate and robust predictions. Our analysis reveals that watch ratio across different duration groups exhibit distinct multi-peak distributions, reflecting the strong aggregation of user behavior. Based on this finding, we employ Vector Quantized Variational Auto-encoder (VQ-VAE) to convert the continuous watch ratio distribution into representative anchors that capture these multi-peak characteristics within each duration group. Subsequently, we project both instance representations and anchors into a common space and utilize Optimal Transport (OT) to generate pseudo-labels aligned with the anchor distribution, allowing instances to obtain structured coordinates within this space during training. Finally, we derive optimized instance representations for watch time prediction by aggregating anchor vectors through weighted integration. Extensive offline experiments on two datasets and large-scale online A/B testing on a short-video platform with over 300 million DAUs demonstrate the consistent superiority of LDA in watch time prediction.

The domain of time series forecasting has gained significant attention due to its critical applications in multimedia-rich web traffic (including video streaming workloads and dynamic content delivery) and cross-platform advertisement click predictions, which are essential for web operations planning. While models like TimeSieve have demonstrated strong capabilities in predicting web visitation metrics, they suffer from critical unfaithfulness issues, including sensitivity to random seeds, input noise, layer noise, and parametric perturbations. To address these limitations, we propose Faithful TimeSieve (FTS), an enhanced framework designed to improve prediction reliability and robustness. Our approach systematically detects and mitigates unfaithfulness in TimeSieve, significantly enhancing its stability and consistency. Experimental results demonstrate that FTS substantially improves the model's faithfulness, setting a new standard for temporal forecasting methods. This advancement not only increases TimeSieve's reliability but also contributes to more robust temporal modeling, particularly crucial for web traffic forecasting where prediction accuracy directly impacts operational decisions. Our work thus represents a significant step toward more dependable time series predictions in web-related applications.

Multi-domain and multi-task learning enhance the efficiency and performance of industrial recommendation systems by integrating information from different domains/tasks to model user interests uniformly. However, existing methods suffer from the problem of representation entanglement, which limits the effective handling of commonality and specificity among various domains/tasks. In this paper, we propose a Prototype-guided Representation Projection (PRP) model to address this issue, which explores a novel direction of applying prototype learning to deal with complex domain/task relationships in the recommendation field. To identify inter-domain/task commonality, PRP initially uses a shared Mixture of Experts (MoE) architecture to learn representations for each sample, projecting them into a common prototype space across all domains/tasks. For domain/task specificity, specific feature extraction experts are employed, and sample representations are projected to the corresponding prototype spaces, constrained by an orthogonal loss to ensure the independence of those spaces. Moreover, PRP utilizes Optimal Transport (OT) to guide the correct representation projection within the prototype spaces, employing the linear combination of prototypes as the new sample representation. We conduct offline experiments on two open-source datasets and deploy our approach in an online system for A/B testing. Extensive experimental results consistently demonstrate that our approach outperforms existing methods.

In large-scale recommendation systems, multimodal (MM) content is increasingly introduced to enhance the generalization of ID features. The rise of Multimodal Large Language Models (MLLMs) enables the construction of unified user and item representations. However, the semantic distribution gap between MM and ID representations leads to \textit{convergence inconsistency} during joint training: the ID branch converges quickly, while the MM branch requires more epochs, thus limiting overall performance. To address this, we propose a two-stage framework including MM representation learning and joint training optimization. First, we fine-tune the MLLM to generate unified user and item representations, and introduce collaborative signals by post-aligning user ID representations to alleviate semantic differences. Then, we propose an Adaptive Gradient Masking (AGM) training strategy to dynamically regulate parameter updates between ID and MLLM branches. AGM estimates the contribution of each representation with mutual information, and applies non-uniform gradient masking at the sub-network level to balance optimization. We provide theoretical analysis of AGM's effectiveness and further introduce an unbiased variant, AGM*, to enhance training stability. Experiments on offline and online A/B tests validate the effectiveness of our approach in mitigating convergence inconsistency and improving performance.

Personalized medication aims to tailor healthcare to individual patient characteristics. However, the heterogeneity of patient data across healthcare systems presents significant challenges to achieving accurate and effective personalized treatments. Ethical concerns further complicate the aggregation of large volumes of data from diverse institutions. Federated Learning (FL) offers a promising decentralized solution by enabling collaborative model training through the exchange of client models rather than raw data, thus preserving privacy. However, existing FL methods often suffer from retrogression during server aggregation, leading to a decline in model performance in real-world medical FL settings. To address data variability in distributed healthcare systems, we introduce Federated Meta-Learning for Personalized Medication (FedMetaMed), which combines federated learning and meta-learning to create models that adapt to diverse patient data across healthcare systems. The FedMetaMed framework aims to produce superior personalized models for individual clients by addressing these limitations. Specifically, we introduce Cumulative Fourier Aggregation (CFA) at the server to improve stability and effectiveness in global knowledge aggregation. CFA achieves this by gradually integrating client models from low to high frequencies. At the client level, we implement a Collaborative Transfer Optimization (CTO) strategy with a three-step process—Retrieve, Reciprocate, and Refine—to enhance the personalized local model through seamless global knowledge transfer. Experiments on real-world medical imaging datasets demonstrate that FedMetaMed outperforms state-of-the-art FL methods, showing superior generalization even on out-of-distribution cohorts.

Automatic Medical Imaging Narrative generation aims to alleviate the workload of radiologists by producing accurate clinical descriptions directly from radiological images. However, the subtle visual nuances and domain-specific terminology in medical images pose significant challenges compared to generic image captioning tasks. Existing approaches often neglect the vital distinction between normal and abnormal findings, leading to suboptimal performance. In this work, we propose FODA-PG, a novel Fine-grained Organ-Disease Adaptive Partitioning Graph framework that addresses these limitations through domain-adaptive learning. FODA-PG constructs a granular graphical representation of radiological findings by separating disease-related attributes into distinct "disease-specific" and "disease-free" categories based on their clinical significance and location. This adaptive partitioning enables our model to capture the nuanced differences between normal and pathological states, mitigating the impact of data biases. By integrating this fine-grained semantic knowledge into a powerful transformer-based architecture and providing rigorous mathematical justifications for its effectiveness, FODA-PG generates precise and clinically coherent reports with enhanced generalization capabilities. Extensive experiments on the IU-Xray and MIMIC-CXR benchmarks demonstrate the superiority of our approach over state-of-the-art methods, highlighting the importance of domain adaptation in medical report generation.

In Internet of Vehicles (IoV), intelligent transportation recognition is key to smart transportation systems. However, training models using data from individual vehicles often results in unreliable performance due to limited storage, leading to new data overwriting old samples. This approach suffers from catastrophic forgetting, low accuracy, and poor generalization. Current industry solutions, such as federated learning, incremental learning (iCaRL), and elastic weight consolidation (EWC), face limitations when dealing with the frequent, dynamic changes in IoV data and the heterogeneity of system structures.In this paper, we propose the FedHotpot framework based on federated learning to tackle these challenges. On the one hand, it employs a dual knowledge distillation strategy to ensure that knowledge learned by the iteratively updated model is not forgotten. On the other hand, we have optimized the model aggregation algorithm to mitigate the adverse effects of system and statistical heterogeneity arising from the diverse clients in the IoV environment on the federated learning architecture. We conducted multiple sets of comparative experiments on the proposed FedHotpot framework. The experimental results demonstrate that FedHotpot performs exceptionally well across multiple datasets, particularly when dealing with non-independently and identically distributed (non-IID) data, where its performance significantly outperforms traditional federated learning methods. These findings not only validate the effectiveness of the dual knowledge distillation strategy, but also underscore the pivotal role of optimized aggregation algorithms in enhancing the overall performance of federated learning.

Co-Speech gesture generation encounters challenges with imbalanced, long-tailed gesture distributions. While recent methods typically address this by employing Vector Quantized Variational Autoencoder (VQ-VAE), encode gestures into a codebook and classify codebook indices based on audio or text cues. However, due to the imbalanced, the codebook classification tends to bias towards majority gestures, neglecting semantically rich minority gestures. To address this, this paper proposes the Entropy-Guided Co-Speech Gesture Generation (EGGesture). EGGesture leverages an Entropy-Guided VQ-VAE to jointly optimizes the distribution of codebook indices and adjusts loss weights for codebook index classification, which consists of a) A differentiable approach for entropy computation using Gumbel-Softmax and cosine similarity, facilitating online codebook distribution optimization, and b) a strategy that utilizes computed codebook entropy to collaboratively guide the classification loss weighting. These designs enable the dynamic refinement of the codebook utilization, striking a balance between the quality of the learned gesture representation and the accuracy of the classification phase. Experiments on the Trinity and BEAT datasets demonstrate EGGesture's state-of-the-art performance both qualitatively and quantitatively.

The rise of the Internet of Things (IoT) has increased standby energy consumption due to the growing number of smart devices in homes. Existing approaches use real-time energy data and machine learning to identify and minimize standby energy for residential energy management but rely on cloud-based data aggregation and collaborative training due to limited edge device data. However, such an approach incurs extra cloud service costs, risks personal data leakage, and fails to capture residence diversity, resulting in suboptimal energy management performance.

Intelligent Transportation Systems provide ubiquitous communication for the driving users through heterogeneous interconnections. The heterogeneous interconnections are required for uninterrupted resource sharing. Spontaneous resource availability due to vehicle speed and infrastructure connectivity disturb prompt service utilization. In this manuscript, a Permissible Service Selection and Allocation (PSSA) method is proposed to address spontaneous issues in vehicular communication and connection. This method considers vehicle displacement and minimum interconnection factors in accessing a cloud service. Both factors and their balancing impact are analyzed throughout the vehicle’s service requesting interval. In this process, random forest learning is induced to identify the balancing factors’ adjustments. The service access is probed through the active infrastructure based on the balancing factor. The ordering process of the learning intervals provides ease of service selection and allocation. In this process, reallocation is not preferred due to the random displacement of the vehicles. Therefore, the interval dropouts are reduced in both handoff and non-handoff communication scenarios. Further metrics such as service ratio, delay, and connectivity are used in validating the proposed method’s performance.

The Internet of Things (IoT) boom has revolutionized almost every corner of people’s daily lives: healthcare, environment, transportation, manufacturing, supply chain, and so on. With the recent development of sensor and communication technology, IoT artifacts, including smart wearables, cameras, smartwatches, and autonomous systems can accurately measure and perceive their surrounding environment. Continuous sensing generates massive amounts of data and presents challenges for machine learning. Deep learning models (e.g., convolution neural networks and recurrent neural networks) have been extensively employed in solving IoT tasks by learning patterns from multi-modal sensory data. Graph neural networks (GNNs), an emerging and fast-growing family of neural network models, can capture complex interactions within sensor topology and have been demonstrated to achieve state-of-the-art results in numerous IoT learning tasks. In this survey, we present a comprehensive review of recent advances in the application of GNNs to the IoT field, including a deep dive analysis of GNN design in various IoT sensing environments, an overarching list of public data and source codes from the collected publications, and future research directions. To keep track of newly published works, we collect representative papers and their open-source implementations and create a Github repository at GNN4IoT.

Renewable energy supply is a promising solution for datacenters' increasing electricity monetary cost, energy consumption and harmful gas emissions. However, due to the instability of renewable energy, insufficient renewable energy supply may lead to the use of stored energy or brown energy. To handle this problem, in this paper, we propose an instability-resilient renewable energy allocation system. We define a job's service-level-objective (SLO) as the successful running probability by only using supplied renewable energy. The system allocates jobs with the same SLO level to the same physical machine (PM) group, and powers each PM group with renewable energy generators that have probability no less than its SLO to produce the amount no less than its energy demand. We use a deep learning technique to predict the probability of producing the amount no less than each value of each renewable energy source, and predict the energy demands of each PM area. We formulate an optimization problem to match renewable energy resources with different instabilities to different PM groups for supply, and use reinforcement learning method and linear programming method to solve it. We further propose an energy-driven computing resource assignment method, which adjusts the amount of computing resource of each job based on job deadline and failure probability in each PM group, and a failure prediction based energy saving method. Real trace driven experiments show that our methods achieve much lower SLO violations, total energy monetary cost and total carbon emission compared to other methods and the effectiveness of individual methods.

The trend of Internet of Things is bringing in millions of new smart devices into homes to increase the quality of human life. However, this enormous number of new devices have also brings in an increasing energy consumption in standby for awaiting wire-less communication or status change. To reduce standby energy, existing approaches use real-time consumption data and machine learning techniques to identify standby energy, but aggregate data or intermediate model training updates in the cloud to collaboratively perform load forecasting, which could directly or indirectly cause personal data leakage, alongside with significant communication bandwidth and extra cloud service monetary cost. On the other hand, such a global collaborative model yields unsatisfactory en-ergy management performance as they fail to capture the diversity of each residence. In this paper, we propose a privacy-preserved, communication-efficient, personalized and cloud-service-free resi-dential energy management system (EMS) with personalized feder-ated deep reinforcement learning (PFDRL) framework to tackle the standby energy reduction in residential building.

Finance is a particularly challenging playground for deep reinforcement learning. However, establishing high-quality market environments and benchmarks for financial reinforcement learning is challenging due to three major factors, namely, low signal-to-noise ratio of financial data, survivorship bias of historical data, and backtesting overfitting. In this paper, we present an openly accessible FinRL-Meta library that has been actively maintained by the AI4Finance community. First, following a DataOps paradigm, we will provide hundreds of market environments through an automatic data curation pipeline that processes dynamic datasets from real-world markets into gym-style market environments. Second, we reproduce popular papers as stepping stones for users to design new trading strategies. We also deploy the library on cloud platforms so that users can visualize their own results and assess the relative performance via community-wise competitions. Third, FinRL-Meta provides tens of Jupyter/Python demos organized into a curriculum and a documentation website to serve the rapidly growing community. FinRL-Meta is available at: \url{https://github.com/AI4Finance-Foundation/FinRL-Meta}

Recent advancements in deep learning techniques have shown great potential for smart Internet of Things (IoT) applications. However, the edge devices of IoT applications often collect and store only limited data, which is insufficient for training modern deep learning models. Collaborative training methods such as cloud computing and federated learning set steps to build robust models for IoT applications, yet these methods bring the concern of data privacy (e.g., untrusted central server, model inversion). On the other hand, directly applying privacy-preserving techniques such as differential privacy can dramatically degrade the performance of IoT applications. Inspired by the development of model personalization, we aim to design a federated learning framework in a personalized fashion to reduce the accuracy loss caused by privacy-preserving techniques. In this paper, we present PFed-LDP, a private and accurate federated local differential privacy (LDP) framework for IoT sensing data. We first design a dynamic layer sharing mechanism to separate the local model into global layers and personalized layers. Second, we apply LDP noise to the global layers and transmit them to the federated learning framework for aggregation. Third, each local client updates their model with local personalized layers and aggregated global layers to perform IoT tasks. Our experiments on real-world datasets show that we only sacrifice 1.6% of accuracy to achieve privacy-preserving IoT applications. We also observe that our method has the smallest accuracy range, which means we can achieve the best performance for the worst performed client.

Prosanta and Biplab presented a lightweight two-factor authentication scheme for the Internet of Things (IoT) devices based on the physical unclonable function (PUF). Their presented scheme was based on the fuzzy extractor and analyzed various security reasonings, such as mutual authentication, session key agreement, privacy and protection against impersonation, message tampering, and replay attacks. In this article, we present sufficient security analysis to demonstrate that the scheme has various security and privacy issues in its setup and authentication phases. We propose a highly secure and robust authentication protocol based on a public key infrastructure (PKI) digital certificate based on two certificate authorities (CAs) for cloud IoT systems. The proposed authentication method is verified and validated using the Tamarin prover and supported with a detailed security and performance analysis discussion. The scheme security and privacy attributes are compared with other IoT authentication schemes. The analysis has proved that the proposed authentication scheme is more secure and highly reliable as compared to the Prosanta and Biplab authentication scheme.

Internet of Things assisted healthcare services grants reliable clinical diagnosis and analysis by exploiting heterogeneous communication and infrastructure elements. Communication is enabled through point-to-point or cluster-to-point between the users and the diagnosis center. In this process, the complication is the resource sharing and diagnosis swiftness invalidating multiple resources. IoT's open and ubiquitous nature results in proactive resource sharing, resulting in delayed transmissions. This manuscript introduces the Redemptive Resource Sharing and Allocation (R2SA) scheme to address this issue. The available health data is accumulated on a first-come-first-serve basis, and the transmitting infrastructure is selected. In this process, the data-to-capacity of the available infrastructure is identified for non-redemptive resource allocation. The extremity of the capacity and unavailability of the resource is then analyzed for parallel processing and allocation. Therefore, the data accumulation and exchange rely on concurrent sharing and resource allocation processes, deferring a better accumulation ratio. The concurrent redemptive selection and sharing reduces transmission delay, improves resource allocation, and reduces transmission complexity. The entire process is managed for transfer learning, data-to-capacity validation, and concurrent recommendation. The first validation knowledge base remains the same/shared for different data accumulation and sharing intervals.

Internet of Things (IoT) assisted healthcare systems are designed for providing ubiquitous access and recommendations for personal and distributed electronic health services. The heterogeneous IoT platform assists healthcare services with reliable data management through dedicated computing devices. Healthcare services' reliability depends upon the efficient handling of heterogeneous data streams due to variations and errors. A Proportionate Data Analytics (PDA) for heterogeneous healthcare data stream processing is introduced in this manuscript. This analytics method differentiates the data streams based on variations and errors for satisfying the service responses. The classification is streamlined using linear regression for segregating errors from the variations in different time intervals. The time intervals are differentiated recurrently after detecting errors in the stream's variation. This process of differentiation and classification retains a high response ratio for healthcare services through spontaneous regressions. The proposed method's performance is analyzed using the metrics accuracy, identification ratio, delivery, variation factor, and processing time.

Vehicular Ad-Hoc Networks (VANETs) provide roadside communication for improving the ease of driving user information exchange. It interconnects hierarchical infrastructure units and other vehicles for real-time traffic and vehicle management applications. The growth of vehicle and information density requires concord information exchange for application responses. In this article, Vehicle-Consensus Routing Management Scheme (VCRMS) is proposed for achieving fair roadside assistance for driving users. The proposed scheme exploits the surrounding vehicle information for infrastructure selection and traffic management. The infrastructure and vehicle information are analyzed for their similarity using deep learning for extracting monotonous decisions. The current and previous decisions are used for succeeding in information selection and traffic information retrieval. This prevents unnecessary data from being congesting the traffic and vehicle management application during driver assistance. The performance shows that the proposed scheme achieves 10.7% high application response, 15.9% less information delay, and 10.7% less traffic for different vehicle velocities.

A large-scale cloud data center needs to provide high service reliability and availability with low failure occurrence probability. However, current large-scale cloud data centers still face high failure rates due to many reasons such as hardware and software failures, which often result in task and job failures. Such failures can severely reduce the reliability of cloud services and also occupy huge amount of resources to recover the service from failures. Therefore, it is important to predict task or job failures before occurrence with high accuracy to avoid unexpected wastage. Many machine learning and deep learning based methods have been proposed for the task or job failure prediction by analyzing past system message logs and identifying the relationship between the data and the failures. In order to further improve the failure prediction accuracy of the previous machine learning and deep learning based methods, in this article, we propose a failure prediction algorithm based on multi-layer Bidirectional Long Short Term Memory (Bi-LSTM) to identify task and job failures in the cloud. The goal of Bi-LSTM failure prediction algorithm is to predict whether the tasks and jobs are failed or completed. The trace-driven experiments show that our algorithm outperforms other state-of-art prediction methods with 93 percent accuracy and 87 percent for task failure and job failures respectively.

The rapid growth of cloud computing in cloud datacenters in recent decades greatly increases the brown energy consumption in datacenters, and hence significant increase of carbon emission that negatively impacts on the environment as well as the monetary cost. More and more cloud service providers are adopting renewable energy as the energy supply to offset the consumption of brown energy. Meanwhile, an increasing number of renewable energy generators have been built to meet the needs. However, the instability of the renewable energy cannot guarantee the support to the datacenter and the energy competition of different datacenters may lead to datacenter energy outage. In this paper, we focus on the problem of how to match different renewable energy generators to the datacenters from different cloud providers to minimize the carbon emission, monetary cost, and service level objective (SLO) violation due to renewable energy shortage. The challenges here are that the datacenters may compete in energy requesting, the renewable energy generation is not stable and the decision should be made quickly. There have been no previous efforts devoting to this problem. To solve the problem, we first test several machine learning techniques on long-term prediction accuracy on renewable energy generation and energy demand using real traces and identify SARIMA for the prediction. We then propose a multi-agent reinforcement learning based method (MARL) for each datacenter to determine how much renewable energy to request from each generator based on the predicted results. We also propose a deadline guaranteed job postponement method (DGJP) to postpone executing unurgent jobs upon insufficient renewable energy supply. The trace-driven experiments show that MARL outperforms other methods by increasing up to 35% SLO satisfaction ratio, and reducing up to 19% (0.33 billion dollars in 90 days) total monetary cost and 33% total carbon emission, and DGJP further improves the performance.

The emergence of radio frequency (RF) dependent device-free indoor occupancy detection has seen slow acceptance due to its high fragility. Experimentation shows that an RF-dependent occupancy detector initially performs well in the room to be sensed. However, once the physical arrangement of objects changes in the room, the performance of the classifier degrades significantly. To address this issue, we propose BLECS, a Bluetooth-dependent indoor occupancy detection system which can adapt itself in the dynamic environment. BLECS uses a reinforcement learning approach to predict the occupancy of an indoor environment and updates its decision policy by interacting with existing IoT devices and sensors in the room. We tested this system in five different rooms for 520 hours in total, involving four occupants. Results show that, BLECS achieves 21.4% performance improvement in a dynamic environment compared to the state-of-the-art supervised learning algorithm with an average F1 score of 86.52%. This system can also predict occupancy with a maximum 89.23% F1 score in a completely unknown environment with no initial trained model.

The fast-growing trend of Internet of Things (IoT) has provided its users with opportunities to improve user experience such as voice assistants, smart cameras, and home energy management systems. Such smart home applications often require large numbers of diverse training data to accomplish a robust model. As single user may not have enough data to train such a model, users intent to collaboratively train their collected data in order to achieve better performance in such applications, which raise the concern of data privacy protection. Existing approaches for collaborative training need to aggregate data or intermediate model training updates in the cloud to perform load forecasting, which could directly or indirectly cause personal data leakage, alongside with significant communication bandwidth and extra cloud service monetary cost.

As a widely used IT service, more and more companies shift their services to cloud datacenters. It is important for cloud service providers (CSPs) to provide cloud service resources with high elasticity and cost-effectiveness and then achieve good quality of service (QoS) for their clients. However, meeting QoS with cost-effective resource is a challenging problem for CSPs because the workloads of Virtual Machines (VMs) experience variation over time. It is highly necessary to provide an accurate VMs workload prediction method for resource provisioning to efficiently manage cloud resources. In this paper, we first compare the performance of representative state-of-the-art workload prediction methods. We suggest a method to conduct the prediction a certain time before the predicted time point in order to allow sufficient time for task scheduling based on predicted workload. To further improve the prediction accuracy, we introduce a clustering based workload prediction method, which first clusters all the tasks into several categories and then trains a prediction model for each category respectively. The trace-driven experiments based on Google cluster trace demonstrates that our clustering based workload prediction methods outperform other comparison methods and improve the prediction accuracy to around 90% both in CPU and memory.

The size and energy consumption of datacenters have been increasing significantly over the past years. As a result, datacenters' increasing electricity monetary cost, energy consumption and energy harmful gas emissions have become a severe problem. Renewable energy supply is widely seen as a promising solution. However, the instability of renewable energy brings about a new challenge since insufficient energy supply may lead to job running interruptions or failures. Though previous works attempt to more accurately predict the amount of produced renewable energy, due to the instability of its influencing factors (e.g., wind, temperature), sufficient renewable energy supply cannot be always guaranteed. To handle this problem, in this paper, we propose allocating jobs with the same service-level-objective (SLO) level to the same physical machine (PM) group, and power each PM group with renewable energy generators that have probability no less than its SLO to produce the amount no less than its energy demand. It ensures that insufficient renewable energy supply will not lead to SLO violations. We use a deep learning technique to predict the probability of producing amount no less than each value of each renewable energy source and predict the energy demands of each PM area. We formulate an optimization problem: how to match renewable energy resources with different instabilities to different PM groups as energy supply in order to minimize the number of SLO violations (due to interruption from insufficient renewable energy supply), total energy monetary cost and total carbon emission. We then use reinforcement learning method and linear programming method to solve the optimization problem. The real trace driven experiments show that our method can achieve much lower SLO violations, total energy monetary cost and total carbon emission compared to other methods.

A large-scale cloud data center needs to provide high service reliability and availability with low failure occurrence probability. However, current large-scale cloud data centers still face high failure rates due to many reasons such as hardware and software failures, which often result in task and job failures. Such failures can severely reduce the reliability of cloud services and also occupy huge amount of resources to recover the service from failures. Therefore, it is important to predict task or job failures before occurrence with high accuracy to avoid unexpected wastage. Many machine learning and deep learning based methods have been proposed for the task or job failure prediction by analyzing past system message logs and identifying the relationship between the data and the failures. In order to further improve the failure prediction accuracy of the previous machine learning and deep learning based methods, in this paper, we propose a failure prediction algorithm based on multi-layer Bidirectional Long Short Term Memory (Bi-LSTM) to identify task and job failures in the cloud. The goal of Bi-LSTM prediction algorithm is to predict whether the tasks and jobs are failed or completed. The trace-driven experiments show that our algorithm outperforms other state-of-art prediction methods with 93% accuracy and 87% for task failure and job failures respectively.