Pub Date : 2025-03-22DOI: 10.1109/TETCI.2025.3569455
Xiaobo Chen;Kaiyuan Wang;Qiaolin Ye
Spatiotemporal traffic data collected by sensor networks is paramount for intelligent transportation systems, but it often suffers from significant missing values, making accurate recovery a critical challenge. This paper introduces a nonconvex low-tubal-rank tensor completion model with temporal regularization (NLTC-TR) to fill in missing data based on the underlying physical properties of spatiotemporal traffic data, including similarity and periodicity. Our model can capture both the global correlations across different modes of traffic data and the local temporal variations. Specifically, we first present a unified nonconvex model to capture tubal rank along different dimensions while considering the specificity in each mode. The nonconvexity achieved by synergizing logarithm and Schatten-p norm serves as a tight rank approximation. In doing so, the low-rank property of traffic data can be better modeled in the transformed domains. Furthermore, by introducing a temporal regularization, we can make better use of the local variation between adjacent moments. To solve this problem, we propose an efficient iterative algorithm based on the alternating direction method of multipliers (ADMM), where each step can be solved in closed form. Numerical experiments on two real-world traffic datasets with varying missing patterns show that our method outperforms existing algorithms, demonstrating its effectiveness in accurately recovering missing traffic data.
{"title":"Nonconvex Low-Tubal-Rank Tensor Completion With Temporal Regularization for Spatiotemporal Traffic Data Recovery","authors":"Xiaobo Chen;Kaiyuan Wang;Qiaolin Ye","doi":"10.1109/TETCI.2025.3569455","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3569455","url":null,"abstract":"Spatiotemporal traffic data collected by sensor networks is paramount for intelligent transportation systems, but it often suffers from significant missing values, making accurate recovery a critical challenge. This paper introduces a nonconvex low-tubal-rank tensor completion model with temporal regularization (NLTC-TR) to fill in missing data based on the underlying physical properties of spatiotemporal traffic data, including similarity and periodicity. Our model can capture both the global correlations across different modes of traffic data and the local temporal variations. Specifically, we first present a unified nonconvex model to capture tubal rank along different dimensions while considering the specificity in each mode. The nonconvexity achieved by synergizing logarithm and Schatten-p norm serves as a tight rank approximation. In doing so, the low-rank property of traffic data can be better modeled in the transformed domains. Furthermore, by introducing a temporal regularization, we can make better use of the local variation between adjacent moments. To solve this problem, we propose an efficient iterative algorithm based on the alternating direction method of multipliers (ADMM), where each step can be solved in closed form. Numerical experiments on two real-world traffic datasets with varying missing patterns show that our method outperforms existing algorithms, demonstrating its effectiveness in accurately recovering missing traffic data.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 6","pages":"4066-4079"},"PeriodicalIF":5.3,"publicationDate":"2025-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145584684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-21DOI: 10.1109/TETCI.2025.3548803
Maoyu Mao;Chungang Yan;Junli Wang;Jun Yang
The rise of Deepfake technology poses a formidable threat to the credibility of both judicial evidence and intellectual property safeguards. Current methods lack the ability to integrate the texture information of facial features into CNNs, despite the fact that fake contents are subtle and pixel-level. Due to the fixed grid kernel structure, CNNs are limited in their ability to describe detailed fine-grained information, making it challenging to achieve accurate image detection through pixel-level fine-grained features. To mitigate this problem, we propose a Pixel Difference Convolution (PDC) to capture local intrinsic detailed patterns via aggregating both intensity and gradient information. To avoid the redundant feature computations generated by PDC and explicitly enhance the representational power of a standard convolutional kernel, we separate PDC into vertical/horizontal and diagonal parts. Furthermore, we propose an Ensemble Dilated Convolution (EDC) to explore long-range contextual dependencies and further boost performance. We introduce a novel network, Pixel Difference Convolutional Network (PDCNet), which is built with PDC and EDC to expose Deepfake by capturing faint traces of tampering hidden in portrait images. By leveraging PDC and EDC in the information propagation process, PDCNet seamlessly incorporates both local and global pixel differences. Comprehensive experiments are performed on three databases, FF++, Celeb-DF, and DFDC to confirm that our PDCNet outperforms existing approaches. Our approach achieves accuracies of 0.9634, 0.9614, and 0.8819 in FF++, Celeb-DF, and DFDC, respectively.
{"title":"Leveraging Pixel Difference Feature for Deepfake Detection","authors":"Maoyu Mao;Chungang Yan;Junli Wang;Jun Yang","doi":"10.1109/TETCI.2025.3548803","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3548803","url":null,"abstract":"The rise of Deepfake technology poses a formidable threat to the credibility of both judicial evidence and intellectual property safeguards. Current methods lack the ability to integrate the texture information of facial features into CNNs, despite the fact that fake contents are subtle and pixel-level. Due to the fixed grid kernel structure, CNNs are limited in their ability to describe detailed fine-grained information, making it challenging to achieve accurate image detection through pixel-level fine-grained features. To mitigate this problem, we propose a Pixel Difference Convolution (PDC) to capture local intrinsic detailed patterns via aggregating both intensity and gradient information. To avoid the redundant feature computations generated by PDC and explicitly enhance the representational power of a standard convolutional kernel, we separate PDC into vertical/horizontal and diagonal parts. Furthermore, we propose an Ensemble Dilated Convolution (EDC) to explore long-range contextual dependencies and further boost performance. We introduce a novel network, Pixel Difference Convolutional Network (PDCNet), which is built with PDC and EDC to expose Deepfake by capturing faint traces of tampering hidden in portrait images. By leveraging PDC and EDC in the information propagation process, PDCNet seamlessly incorporates both local and global pixel differences. Comprehensive experiments are performed on three databases, FF++, Celeb-DF, and DFDC to confirm that our PDCNet outperforms existing approaches. Our approach achieves accuracies of 0.9634, 0.9614, and 0.8819 in FF++, Celeb-DF, and DFDC, respectively.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 4","pages":"3178-3188"},"PeriodicalIF":5.3,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687698","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-19DOI: 10.1109/TETCI.2025.3547856
Shuo Zhao;Kun Wu;Xin Li;Ying-Chi Chen
Optical neural networks (ONNs) have emerged as a promising solution for energy-efficient deep learning. However, their resource-intensive manufacturing process necessitates efficient methods to streamline ONN architectures without sacrificing their performances. Weight pruning presents a potential remedy. Unlike the conventional neural networks, the pruned weights in ONNs are not necessarily zero in general, thereby making most traditional pruning methods inefficient. In this paper, we propose a novel two-stage pruning method tailored for ONNs. In the first stage, a first-order Taylor expansion of the loss function is applied to effectively identify and prune unimportant weights. To determine the shared value for the pruned weights, a novel optimization method is developed. In the second stage, fine-tuning is further applied to adjust the unpruned weights alongside the shared value of pruned weights. Experimental results on multiple public datasets demonstrate the efficacy of our proposed approach. It achieves superior model compression with minimum loss in accuracy over other conventional pruning techniques.
{"title":"Efficient Weight Pruning for Optical Neural Networks: When Pruned Weights are Non-Zeros","authors":"Shuo Zhao;Kun Wu;Xin Li;Ying-Chi Chen","doi":"10.1109/TETCI.2025.3547856","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3547856","url":null,"abstract":"Optical neural networks (ONNs) have emerged as a promising solution for energy-efficient deep learning. However, their resource-intensive manufacturing process necessitates efficient methods to streamline ONN architectures without sacrificing their performances. Weight pruning presents a potential remedy. Unlike the conventional neural networks, the pruned weights in ONNs are not necessarily zero in general, thereby making most traditional pruning methods inefficient. In this paper, we propose a novel two-stage pruning method tailored for ONNs. In the first stage, a first-order Taylor expansion of the loss function is applied to effectively identify and prune unimportant weights. To determine the shared value for the pruned weights, a novel optimization method is developed. In the second stage, fine-tuning is further applied to adjust the unpruned weights alongside the shared value of pruned weights. Experimental results on multiple public datasets demonstrate the efficacy of our proposed approach. It achieves superior model compression with minimum loss in accuracy over other conventional pruning techniques.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 5","pages":"3572-3581"},"PeriodicalIF":5.3,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145128522","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-18DOI: 10.1109/TETCI.2025.3547851
Shuai Wang;Ting Yu;Shan Pan;Wei Chen;Zehua Wang;Victor C. M. Leung;Zijian Tian
Extracting 3D information from 2D images is highly significant, and self-supervised monocular depth estimation has demonstrated great potential in this field. However, existing methods primarily focus on estimating depth from immediate visual features, leading to severe foreground-background adhesion, which poses challenges for achieving precise depth estimation. In this paper, we propose a depth estimation method called LOEDepth, which can implicitly distinguish foreground objects from the background. In LOEDepth, a latent object embedding module is introduced, which leverages a set of learnable queries to generate latent object proposals from both immediate visual features extracted by the encoder and sparse object features derived through multi-scale deformable attention. These latent object proposals are utilized to perform soft classification on the decoded features to distinguish foreground objects from the background. Additionally, as depth boundaries do not always align with semantic boundaries, we propose a novel deep decoder to provide decoding features with rich spatial location retrieval and semantic information. Finally, two mask strategies are utilized to conceal pixels violating the scene's static assumption, so as to mitigate disruptions caused by abnormal pixels during self-supervised training. Experimental results on the KITTI and Make3D datasets demonstrate significant performance improvements and robust fine-grained scene depth estimation capabilities of the proposed method.
{"title":"Latent Object Embedding for Self-Supervised Monocular Depth Estimation","authors":"Shuai Wang;Ting Yu;Shan Pan;Wei Chen;Zehua Wang;Victor C. M. Leung;Zijian Tian","doi":"10.1109/TETCI.2025.3547851","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3547851","url":null,"abstract":"Extracting 3D information from 2D images is highly significant, and self-supervised monocular depth estimation has demonstrated great potential in this field. However, existing methods primarily focus on estimating depth from immediate visual features, leading to severe foreground-background adhesion, which poses challenges for achieving precise depth estimation. In this paper, we propose a depth estimation method called LOEDepth, which can implicitly distinguish foreground objects from the background. In LOEDepth, a latent object embedding module is introduced, which leverages a set of learnable queries to generate latent object proposals from both immediate visual features extracted by the encoder and sparse object features derived through multi-scale deformable attention. These latent object proposals are utilized to perform soft classification on the decoded features to distinguish foreground objects from the background. Additionally, as depth boundaries do not always align with semantic boundaries, we propose a novel deep decoder to provide decoding features with rich spatial location retrieval and semantic information. Finally, two mask strategies are utilized to conceal pixels violating the scene's static assumption, so as to mitigate disruptions caused by abnormal pixels during self-supervised training. Experimental results on the KITTI and Make3D datasets demonstrate significant performance improvements and robust fine-grained scene depth estimation capabilities of the proposed method.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 5","pages":"3548-3559"},"PeriodicalIF":5.3,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145128552","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the realm of multi-goal reinforcement learning for robot manipulation, effectively addressing sparse rewards has been a key challenge. The hindsight experience replay (HER) mechanism has provided notable advancements in this domain, yet its efficiency and adaptability still require further improvement. This paper introduces TUCA-HER for robot manipulation skill learning via Trajectory Utility and Conservative Advantage. We start by computing trajectory utility for experience samples collected in the early stages of training, which allows for dynamic relabeling and significantly enhances sample efficiency. Furthermore, we integrate conservative advantage learning into the actor-critic framework, reshaping rewards to construct TUCA-HER. Finally, we apply TUCA-HER to robot manipulation skill learning tasks, providing details on algorithmic implementation and complexity analysis. Evaluations conducted on OpenAI Fetch and Hand environments demonstrate TUCA-HER's superior performance in sample efficiency and task success rate compared to other algorithms. Notably, in the FetchPickAndPlace task, TUCA-HER showcases a remarkable 46% improvement over the Double experience replay buffer Adaptive Soft Hindsight Experience Replay (DAS-HER). Furthermore, Sim-to-Real experiments are conducted to validate the effectiveness of TUCA-HER in real-world environments.
{"title":"TUCA-HER: An Improved HER for Robot Manipulation Skill Learning via Trajectory Utility and Conservative Advantage","authors":"Peiliang Wu;Zhaoqi Wang;Yao Li;Wenbai Chen;Guowei Gao","doi":"10.1109/TETCI.2025.3548787","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3548787","url":null,"abstract":"In the realm of multi-goal reinforcement learning for robot manipulation, effectively addressing sparse rewards has been a key challenge. The hindsight experience replay (HER) mechanism has provided notable advancements in this domain, yet its efficiency and adaptability still require further improvement. This paper introduces TUCA-HER for robot manipulation skill learning via Trajectory Utility and Conservative Advantage. We start by computing trajectory utility for experience samples collected in the early stages of training, which allows for dynamic relabeling and significantly enhances sample efficiency. Furthermore, we integrate conservative advantage learning into the actor-critic framework, reshaping rewards to construct TUCA-HER. Finally, we apply TUCA-HER to robot manipulation skill learning tasks, providing details on algorithmic implementation and complexity analysis. Evaluations conducted on OpenAI Fetch and Hand environments demonstrate TUCA-HER's superior performance in sample efficiency and task success rate compared to other algorithms. Notably, in the FetchPickAndPlace task, TUCA-HER showcases a remarkable 46% improvement over the Double experience replay buffer Adaptive Soft Hindsight Experience Replay (DAS-HER). Furthermore, Sim-to-Real experiments are conducted to validate the effectiveness of TUCA-HER in real-world environments.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 5","pages":"3560-3571"},"PeriodicalIF":5.3,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145128539","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-18DOI: 10.1109/TETCI.2025.3547635
Tao Wang;Xinlin Zhang;Yuanbo Zhou;Yuanbin Chen;Longxuan Zhao;Tao Tan;Tong Tong
In recent years, supervised learning using convolutional neural networks (CNN) has served as a benchmark for various medical image segmentation and classification. However, supervised learning deeply relies on large-scale annotated data, which is expensive, time-consuming, and even impractical to acquire in medical imaging applications. Moreover, effective utilization of annotation resources might not always be feasible during the annotation process. To optimize the utilization of annotation resources, a proposed active learning framework is introduced that is applicable to both 2D and 3D segmentation and classification tasks. This framework aims to reduce annotation costs by selecting more valuable samples for annotation from the pool of unlabeled data. Based on the perturbation consistency, we apply different perturbations to the input data and propose a perturbation consistency evaluation module to evaluate the consistency among predictions when applying different perturbations to the data. Subsequently, we rank the consistency of each data and select samples with lower consistency as high-value candidates. These selected samples are prioritized for annotation. We extensively validated our proposed framework on three publicly available and challenging medical image datasets, Kvasir Dataset, COVID-19 Infection Segmentation Dataset, and BraTS2019 Dataset. The experimental results demonstrate that our proposed framework can achieve significantly improved performance with fewer annotations in 2D classification and segmentation and 3D segmentation tasks. The proposed framework enables more efficient utilization of annotation resources by annotating more representative samples, thus enhancing the model's robustness with fewer annotation costs.
{"title":"PCDAL: A Perturbation Consistency-Driven Active Learning Approach for Medical Image Segmentation and Classification","authors":"Tao Wang;Xinlin Zhang;Yuanbo Zhou;Yuanbin Chen;Longxuan Zhao;Tao Tan;Tong Tong","doi":"10.1109/TETCI.2025.3547635","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3547635","url":null,"abstract":"In recent years, supervised learning using convolutional neural networks (CNN) has served as a benchmark for various medical image segmentation and classification. However, supervised learning deeply relies on large-scale annotated data, which is expensive, time-consuming, and even impractical to acquire in medical imaging applications. Moreover, effective utilization of annotation resources might not always be feasible during the annotation process. To optimize the utilization of annotation resources, a proposed active learning framework is introduced that is applicable to both 2D and 3D segmentation and classification tasks. This framework aims to reduce annotation costs by selecting more valuable samples for annotation from the pool of unlabeled data. Based on the perturbation consistency, we apply different perturbations to the input data and propose a perturbation consistency evaluation module to evaluate the consistency among predictions when applying different perturbations to the data. Subsequently, we rank the consistency of each data and select samples with lower consistency as high-value candidates. These selected samples are prioritized for annotation. We extensively validated our proposed framework on three publicly available and challenging medical image datasets, Kvasir Dataset, COVID-19 Infection Segmentation Dataset, and BraTS2019 Dataset. The experimental results demonstrate that our proposed framework can achieve significantly improved performance with fewer annotations in 2D classification and segmentation and 3D segmentation tasks. The proposed framework enables more efficient utilization of annotation resources by annotating more representative samples, thus enhancing the model's robustness with fewer annotation costs.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 4","pages":"3162-3177"},"PeriodicalIF":5.3,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-17DOI: 10.1109/TETCI.2025.3540420
Yanli Ren;Zhe Yang;Guorui Feng;Xinpeng Zhang
Privacy-preserving federated learning (PPFL) allows multiple users to collaboratively train models on local devices without the the risk of privacy leakage. However, PPFL is prone to be disrupted by poisoning attacks for the server being forbbiden from accessing users' updates. The existing protocols focusing on poisoning attacks in PPFL generally use two servers to interactively execute protocols to defend against poisoning attacks, while the other ones using a single server require multiple rounds of server-user interactions, both of which incur significant communication overheads. We propose PurifyFL, a privacy-preserving poisoning attacks defense strategy. PurifyFL only relies on a single server while most of the previous works depend on two non-colluding servers, which are impractical in reality. Moreover, We also achieve non-interactivity between the users and the server. Experiments show that PurifyFL can effectively resist typical poisoning attacks with lower computational and communication overheads compared to existing works.
{"title":"PurifyFL: Non-Interactive Privacy-Preserving Federated Learning Against Poisoning Attacks Based on Single Server","authors":"Yanli Ren;Zhe Yang;Guorui Feng;Xinpeng Zhang","doi":"10.1109/TETCI.2025.3540420","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3540420","url":null,"abstract":"Privacy-preserving federated learning (PPFL) allows multiple users to collaboratively train models on local devices without the the risk of privacy leakage. However, PPFL is prone to be disrupted by poisoning attacks for the server being forbbiden from accessing users' updates. The existing protocols focusing on poisoning attacks in PPFL generally use two servers to interactively execute protocols to defend against poisoning attacks, while the other ones using a single server require multiple rounds of server-user interactions, both of which incur significant communication overheads. We propose PurifyFL, a privacy-preserving poisoning attacks defense strategy. PurifyFL only relies on a single server while most of the previous works depend on two non-colluding servers, which are impractical in reality. Moreover, We also achieve non-interactivity between the users and the server. Experiments show that PurifyFL can effectively resist typical poisoning attacks with lower computational and communication overheads compared to existing works.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 3","pages":"2232-2243"},"PeriodicalIF":5.3,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144148174","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Automated guided vehicles are widely utilized in the real production environment for tasks such as job transfer and inter-factory collaboration, yet they remain relatively underexplored in academic research. This study addresses the distributed permutation flow shop co-scheduling problem with sequence-dependent setup times (DPFCSP-SDST). We propose a novel solution that leverages an optimization algorithm, specifically a learning-based two-stage multi-thread iterated greedy algorithm (LTMIG). First, a problem-specific initialization method is designed to generate the initialization solution in two stages. Second, a Q-learning-based operator adaptation strategy is adopted to guide the evolutionary direction of factory assignment to reduce the makespan. Then, the proposed destructive-construction strategy builds an archive set to share historical knowledge with different stages of search, ensuring exploration capability. Local search effectively combines the parallel computing power of multi-threading with the inherent exploitation capability of LTMIG, and fully utilizes the information of elite solutions. Extensive experimental results demonstrate that LTMIG is significantly better than the comparison algorithms mentioned in the paper, and it turns out that LTMIG is the most suitable algorithm for solving DPFCSP-SDST.
{"title":"A Learning-Based Two-Stage Multi-Thread Iterated Greedy Algorithm for Co-Scheduling of Distributed Factories and Automated Guided Vehicles With Sequence-Dependent Setup Times","authors":"Zijiang Liu;Hongyan Sang;Biao Zhang;Leilei Meng;Tao Meng","doi":"10.1109/TETCI.2025.3540405","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3540405","url":null,"abstract":"Automated guided vehicles are widely utilized in the real production environment for tasks such as job transfer and inter-factory collaboration, yet they remain relatively underexplored in academic research. This study addresses the distributed permutation flow shop co-scheduling problem with sequence-dependent setup times (DPFCSP-SDST). We propose a novel solution that leverages an optimization algorithm, specifically a learning-based two-stage multi-thread iterated greedy algorithm (LTMIG). First, a problem-specific initialization method is designed to generate the initialization solution in two stages. Second, a Q-learning-based operator adaptation strategy is adopted to guide the evolutionary direction of factory assignment to reduce the makespan. Then, the proposed destructive-construction strategy builds an archive set to share historical knowledge with different stages of search, ensuring exploration capability. Local search effectively combines the parallel computing power of multi-threading with the inherent exploitation capability of LTMIG, and fully utilizes the information of elite solutions. Extensive experimental results demonstrate that LTMIG is significantly better than the comparison algorithms mentioned in the paper, and it turns out that LTMIG is the most suitable algorithm for solving DPFCSP-SDST.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 3","pages":"2208-2218"},"PeriodicalIF":5.3,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144148133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Heterogeneous Graph Neural Networks (HetGNN) have garnered significant attention and demonstrated success in tackling various tasks. However, most existing HetGNNs face challenges in effectively addressing unreliable heterogeneous graph structures and encounter semantic indistinguishability problems as their depth increases. In an effort to deal with these challenges, we introduce a novel heterogeneous graph representation learning with optimized structures to optimize heterogeneous graph structures and utilize semantic aggregation mechanism to alleviate semantic indistinguishability while learning node embeddings. To address the heterogeneity of relations within heterogeneous graphs, the proposed algorithm employs a strategy of generating distinct relational subgraphs and incorporating them with node features to optimize structural learning. To resolve the issue of semantic indistinguishability, the proposed algorithm adopts a semantic aggregation mechanism to assign appropriate weights to different meta-paths, consequently enhancing the effectiveness of captured node features. This methodology enables the learning of distinguishable node embeddings by a deeper HetGNN model. Extensive experiments on the node classification task validate the promising performance of the proposed framework when compared with state-of-the-art methods.
{"title":"HGRL-S: Towards Heterogeneous Graph Representation Learning With Optimized Structures","authors":"Shanfeng Wang;Dong Wang;Xiaona Ruan;Xiaolong Fan;Maoguo Gong;He Zhang","doi":"10.1109/TETCI.2025.3543414","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3543414","url":null,"abstract":"Heterogeneous Graph Neural Networks (HetGNN) have garnered significant attention and demonstrated success in tackling various tasks. However, most existing HetGNNs face challenges in effectively addressing unreliable heterogeneous graph structures and encounter semantic indistinguishability problems as their depth increases. In an effort to deal with these challenges, we introduce a novel heterogeneous graph representation learning with optimized structures to optimize heterogeneous graph structures and utilize semantic aggregation mechanism to alleviate semantic indistinguishability while learning node embeddings. To address the heterogeneity of relations within heterogeneous graphs, the proposed algorithm employs a strategy of generating distinct relational subgraphs and incorporating them with node features to optimize structural learning. To resolve the issue of semantic indistinguishability, the proposed algorithm adopts a semantic aggregation mechanism to assign appropriate weights to different meta-paths, consequently enhancing the effectiveness of captured node features. This methodology enables the learning of distinguishable node embeddings by a deeper HetGNN model. Extensive experiments on the node classification task validate the promising performance of the proposed framework when compared with state-of-the-art methods.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 3","pages":"2359-2370"},"PeriodicalIF":5.3,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144148146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-12DOI: 10.1109/TETCI.2025.3543769
Wei Dai;Teng Cui;Tong Zhang;Badong Chen
Vertical Federated Learning (VFL) enables the construction of models by combining clients with different features without compromising privacy. Existing VFL methods exhibit tightly coupled participant parameters, resulting in substantial interdependencies among clients during the prediction phase, which significantly hampers the model's usability. To tackle these challenges, this paper studies a VFL approach with low coupling of parameters between clients. Drawing inspiration from federated cooperation and teacher-supervised learning, we propose a low-coupling vertical federated learning with privileged information paradigm (VFL+), allowing participants to make autonomous predictions. Specifically, VFL+ treats information from other clients as privileged data during the training phase rather than the testing phase, thereby achieving independence in individual model predictions. Subsequently, this paper further investigates three typical scenarios of vertical cooperation and designs corresponding cooperative frameworks. Systematic experiments on real data sets demonstrate the effectiveness of the proposed method.
{"title":"VFL+: Low-Coupling Vertical Federated Learning With Privileged Information Paradigm","authors":"Wei Dai;Teng Cui;Tong Zhang;Badong Chen","doi":"10.1109/TETCI.2025.3543769","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3543769","url":null,"abstract":"Vertical Federated Learning (VFL) enables the construction of models by combining clients with different features without compromising privacy. Existing VFL methods exhibit tightly coupled participant parameters, resulting in substantial interdependencies among clients during the prediction phase, which significantly hampers the model's usability. To tackle these challenges, this paper studies a VFL approach with low coupling of parameters between clients. Drawing inspiration from federated cooperation and teacher-supervised learning, we propose a low-coupling vertical federated learning with privileged information paradigm (VFL+), allowing participants to make autonomous predictions. Specifically, VFL+ treats information from other clients as privileged data during the training phase rather than the testing phase, thereby achieving independence in individual model predictions. Subsequently, this paper further investigates three typical scenarios of vertical cooperation and designs corresponding cooperative frameworks. Systematic experiments on real data sets demonstrate the effectiveness of the proposed method.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 5","pages":"3533-3547"},"PeriodicalIF":5.3,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145128536","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：