Pub Date : 2026-05-01Epub Date: 2026-02-14DOI: 10.1016/j.neucom.2026.133086
Wei Cao , Shanshan Wang
Expectile regression neural network (ERNN) is powerful tool for capturing heterogeneity and complex nonlinear structures in data. However, most existing research has primarily focused on fully observed data, with reletive limited attention given to scenarios involving censored observations. In this paper, we propose a data-augmentation–based ERNN algorithm, termed DAERNN, for modeling heterogeneous censored data with complex relationships among variables. The proposed DAERNN is flexible and capable of exploring potential nonlinear effects of covariates on the conditional expectiles of the response under various types of censoring, thereby enhancing its applicability to practical censored data analysis. Moreover, DAERNN can be readily implemented via a data augmentation strategy combined with a standard gradient-based optimization algorithm, which directly yields estimates of the conditional expectile functions. The advantages of the DAERNN are illustrated through extensive Monte Carlo simulation studies and two real data applications. The results show that DAERNN outperforms existing censored ERNN methods as well as other competing approaches.
{"title":"Neural network for censored expectile regression based on data augmentation","authors":"Wei Cao , Shanshan Wang","doi":"10.1016/j.neucom.2026.133086","DOIUrl":"10.1016/j.neucom.2026.133086","url":null,"abstract":"<div><div>Expectile regression neural network (ERNN) is powerful tool for capturing heterogeneity and complex nonlinear structures in data. However, most existing research has primarily focused on fully observed data, with reletive limited attention given to scenarios involving censored observations. In this paper, we propose a data-augmentation–based ERNN algorithm, termed DAERNN, for modeling heterogeneous censored data with complex relationships among variables. The proposed DAERNN is flexible and capable of exploring potential nonlinear effects of covariates on the conditional expectiles of the response under various types of censoring, thereby enhancing its applicability to practical censored data analysis. Moreover, DAERNN can be readily implemented via a data augmentation strategy combined with a standard gradient-based optimization algorithm, which directly yields estimates of the conditional expectile functions. The advantages of the DAERNN are illustrated through extensive Monte Carlo simulation studies and two real data applications. The results show that DAERNN outperforms existing censored ERNN methods as well as other competing approaches.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 133086"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2026-05-01Epub Date: 2026-02-11DOI: 10.1016/j.neucom.2026.133014
ZhuoLin Li , NingNing Cui , Huo Wu , Zhongyun Bao , Subin Huang
Accurate spatiotemporal forecasting hinges on capturing the intricate and dynamic relationships within data. While hypergraph neural networks have shown promise for capturing high-order interactions, existing methods typically rely on external prior knowledge or learned static hypergraph structures, thereby limiting their ability to capture dynamic variations. Moreover, directly learning the full hypergraph incidence matrix suffers from parameter redundancy and overfitting. To mitigate these drawbacks, we propose a novel Dynamic Hypergraph and Graph Structure Inference Model (DHGSIM), which simultaneously models pairwise and high-order relationships without external prior knowledge. Specifically, for dynamic high-order associations, we construct hypergraph structures leveraging low-rank factorization to boost parameter efficiency and mitigate overfitting. A dynamic routing mechanism is further applied to enable the learned hypergraph structure to interact adaptively with the input data, thereby refining hyperedge representations. In pairwise association modeling, we propose a dynamic graph structure learning method that incorporates a key node identification mechanism to capture crucial interactions. Finally, we decouple temporal and spatial feature extraction to improve efficiency and optimize the entire framework end-to-end. Comprehensive experiments on five widely-used benchmark datasets show that our method attains superior performance. The source code is publicly available at https://github.com/ZhuoLinLi-shu/DHGSIM.
{"title":"Dynamic hypergraph structure learning for spatio-temporal time series forecasting","authors":"ZhuoLin Li , NingNing Cui , Huo Wu , Zhongyun Bao , Subin Huang","doi":"10.1016/j.neucom.2026.133014","DOIUrl":"10.1016/j.neucom.2026.133014","url":null,"abstract":"<div><div>Accurate spatiotemporal forecasting hinges on capturing the intricate and dynamic relationships within data. While hypergraph neural networks have shown promise for capturing high-order interactions, existing methods typically rely on external prior knowledge or learned static hypergraph structures, thereby limiting their ability to capture dynamic variations. Moreover, directly learning the full hypergraph incidence matrix suffers from parameter redundancy and overfitting. To mitigate these drawbacks, we propose a novel Dynamic Hypergraph and Graph Structure Inference Model (DHGSIM), which simultaneously models pairwise and high-order relationships without external prior knowledge. Specifically, for dynamic high-order associations, we construct hypergraph structures leveraging low-rank factorization to boost parameter efficiency and mitigate overfitting. A dynamic routing mechanism is further applied to enable the learned hypergraph structure to interact adaptively with the input data, thereby refining hyperedge representations. In pairwise association modeling, we propose a dynamic graph structure learning method that incorporates a key node identification mechanism to capture crucial interactions. Finally, we decouple temporal and spatial feature extraction to improve efficiency and optimize the entire framework end-to-end. Comprehensive experiments on five widely-used benchmark datasets show that our method attains superior performance. The source code is publicly available at <span><span>https://github.com/ZhuoLinLi-shu/DHGSIM</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 133014"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172688","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2026-05-01Epub Date: 2026-02-07DOI: 10.1016/j.neucom.2026.132990
Yue Zhao, Zhenjiao Lin, Feng Qi
Spiking Neural P systems (SN P systems), as a member of the third generation artificial neural network models, have been widely studied in recent years due to their high scalability, high parallelism, and low energy consumption. However, the original SN P systems have limitations in nonlinear learning and dynamic plasticity. We introduced brain-derived neurotrophic factor (BDNF) into SN P systems for the first time, integrated BDNF and its signaling pathways, and enhanced spike control and synaptic plasticity through four innovative rules, thereby simulating the role of BDNF in neuronal adaptation. We demonstrate that the system requires only 25 neurons to perform universal calculations, which is less than the number of neurons used by existing SN P variants. Additionally, experimental evaluations on function approximation and image classification tasks confirm that the model achieves state-of-the-art performance. In the function fitting task, the system structure is further simplified through visual training and pruning strategies while maintaining efficient nonlinear computing power and high interpretability, fully demonstrating the operation process of the BDNF-SN P systems in actual tasks. In image classification, both comparative and ablation studies across MNIST and four MedMNIST datasets confirm the superiority of the 5B-BDNF-SN P family, with the 5B-Gram-SN P variant achieving up to 99.56% accuracy on MNIST and 97.87% AUC on PathMNIST. Furthermore, the model demonstrates high inference efficiency with fewer parameters, validating its effectiveness and adaptability in both general and medical image classification tasks.
{"title":"Spiking neural P systems with brain-derived neurotrophic factor","authors":"Yue Zhao, Zhenjiao Lin, Feng Qi","doi":"10.1016/j.neucom.2026.132990","DOIUrl":"10.1016/j.neucom.2026.132990","url":null,"abstract":"<div><div>Spiking Neural P systems (SN P systems), as a member of the third generation artificial neural network models, have been widely studied in recent years due to their high scalability, high parallelism, and low energy consumption. However, the original SN P systems have limitations in nonlinear learning and dynamic plasticity. We introduced brain-derived neurotrophic factor (BDNF) into SN P systems for the first time, integrated BDNF and its signaling pathways, and enhanced spike control and synaptic plasticity through four innovative rules, thereby simulating the role of BDNF in neuronal adaptation. We demonstrate that the system requires only 25 neurons to perform universal calculations, which is less than the number of neurons used by existing SN P variants. Additionally, experimental evaluations on function approximation and image classification tasks confirm that the model achieves state-of-the-art performance. In the function fitting task, the system structure is further simplified through visual training and pruning strategies while maintaining efficient nonlinear computing power and high interpretability, fully demonstrating the operation process of the BDNF-SN P systems in actual tasks. In image classification, both comparative and ablation studies across MNIST and four MedMNIST datasets confirm the superiority of the 5B-BDNF-SN P family, with the 5B-Gram-SN P variant achieving up to 99.56% accuracy on MNIST and 97.87% AUC on PathMNIST. Furthermore, the model demonstrates high inference efficiency with fewer parameters, validating its effectiveness and adaptability in both general and medical image classification tasks.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 132990"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2026-05-01Epub Date: 2026-02-11DOI: 10.1016/j.neucom.2026.133032
Haoran Li , Zhiqiang Lv , Zhaobin Ma , Jianbo Li , Xiaolong Ma , Dongxin Sun , Kangxin Guo , Jun Liu
The Intelligent Transport Systems represent a pivotal research area within the broader context of smart city construction. It constitutes a vital component of the contemporary urban transport system, with the potential to facilitate high-quality development. The prediction of traffic flow represents a significant research area within the field of ITS. It offers a valuable opportunity to develop a robust data foundation for the planning and optimisation of urban traffic road networks. The majority of studies in this field currently employ static graphs and graph neural networks to complete the traffic flow prediction task. The use of static graphs for traffic flow prediction is inadequate for capturing the dynamic spatial and temporal characteristics of the traffic network structure. Furthermore, graph neural networks are only capable of performing local spatial characteristic analysis. To address the issue of global feature analysis of traffic network topology, multi-layer graph neural networks are required for iterative computation. The number of layers of graph neural networks increases in line with the number of nodes in the traffic network. To address the aforementioned issues, this study proposes a neural network architecture that employs a tree structure for attention computation, namely the Spatio-temporal Tree Attention Network (STTAT). In particular, this study proposes a tree-structured representation of traffic network topology and a tree-structured attention computation method for learning global features of traffic network topology. The proposed model has been evaluated on several real-world traffic datasets, and its performance has been compared with that of several baseline models. The results demonstrate that the proposed model significantly outperforms the baseline models in terms of prediction accuracy.
在智慧城市建设的大背景下，智能交通系统是一个关键的研究领域。它是当代城市交通系统的重要组成部分，具有促进高质量发展的潜力。交通流预测是智能交通领域的一个重要研究领域。它提供了一个宝贵的机会，为规划和优化城市交通道路网络建立一个强大的数据基础。目前该领域的研究大多采用静态图和图神经网络来完成交通流预测任务。使用静态图形进行交通流预测不足以捕捉交通网络结构的动态时空特征。此外，图神经网络只能进行局部空间特征分析。为了解决交通网络拓扑结构的全局特征分析问题，需要使用多层图神经网络进行迭代计算。图神经网络的层数随交通网络中节点数的增加而增加。针对上述问题，本研究提出了一种采用树状结构进行注意力计算的神经网络架构，即时空树状注意力网络（spatial -temporal tree attention network， STTAT）。特别地，本研究提出了一种树形的交通网络拓扑表示和一种树形的注意力计算方法来学习交通网络拓扑的全局特征。该模型在多个真实交通数据集上进行了评估，并与多个基线模型的性能进行了比较。结果表明，该模型在预测精度上明显优于基线模型。
{"title":"Spatio-temporal tree attention network for forecasting traffic flow","authors":"Haoran Li , Zhiqiang Lv , Zhaobin Ma , Jianbo Li , Xiaolong Ma , Dongxin Sun , Kangxin Guo , Jun Liu","doi":"10.1016/j.neucom.2026.133032","DOIUrl":"10.1016/j.neucom.2026.133032","url":null,"abstract":"<div><div>The Intelligent Transport Systems represent a pivotal research area within the broader context of smart city construction. It constitutes a vital component of the contemporary urban transport system, with the potential to facilitate high-quality development. The prediction of traffic flow represents a significant research area within the field of ITS. It offers a valuable opportunity to develop a robust data foundation for the planning and optimisation of urban traffic road networks. The majority of studies in this field currently employ static graphs and graph neural networks to complete the traffic flow prediction task. The use of static graphs for traffic flow prediction is inadequate for capturing the dynamic spatial and temporal characteristics of the traffic network structure. Furthermore, graph neural networks are only capable of performing local spatial characteristic analysis. To address the issue of global feature analysis of traffic network topology, multi-layer graph neural networks are required for iterative computation. The number of layers of graph neural networks increases in line with the number of nodes in the traffic network. To address the aforementioned issues, this study proposes a neural network architecture that employs a tree structure for attention computation, namely the Spatio-temporal Tree Attention Network (STTAT). In particular, this study proposes a tree-structured representation of traffic network topology and a tree-structured attention computation method for learning global features of traffic network topology. The proposed model has been evaluated on several real-world traffic datasets, and its performance has been compared with that of several baseline models. The results demonstrate that the proposed model significantly outperforms the baseline models in terms of prediction accuracy.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 133032"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386719","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Time series forecasting plays a crucial role in decision-making across various scenarios, including weather, traffic, and finance. Real-world time series typically exhibit complex patterns where multiple periods and dynamic trends overlap. However, traditional models based on fixed receptive fields struggle to adapt flexibly to scale variations in periodic and trend features across different application scenarios, thereby limiting their generalization capabilities. In this paper, we propose a scale-aware Adaptive Receptive Field Network (ARFNet). Specifically, we propose an Adaptive Receptive Field Pyramid module that dynamically adjusts the receptive field size based on the inherent characteristics of the data, providing high-quality periodic and trend feature inputs for subsequent analysis. In addition, we design a Scale-Aware Feature Synthesizer module to capture the dependencies between features at different scales, enhancing the model’s comprehension and utilization of multi-scale temporal information. Extensive experimental results on real-world datasets show that ARFNet outperforms state-of-the-art methods in time series forecasting.
{"title":"ARFNet: Scale-aware adaptive receptive field network for time series forecasting","authors":"Chongyun Qin, Zhenpeng Wu, Yiting Shi, Jianliang Gao","doi":"10.1016/j.neucom.2026.133044","DOIUrl":"10.1016/j.neucom.2026.133044","url":null,"abstract":"<div><div>Time series forecasting plays a crucial role in decision-making across various scenarios, including weather, traffic, and finance. Real-world time series typically exhibit complex patterns where multiple periods and dynamic trends overlap. However, traditional models based on fixed receptive fields struggle to adapt flexibly to scale variations in periodic and trend features across different application scenarios, thereby limiting their generalization capabilities. In this paper, we propose a scale-aware <strong>A</strong>daptive <strong>R</strong>eceptive <strong>F</strong>ield <strong>Net</strong>work (<strong>ARFNet</strong>). Specifically, we propose an Adaptive Receptive Field Pyramid module that dynamically adjusts the receptive field size based on the inherent characteristics of the data, providing high-quality periodic and trend feature inputs for subsequent analysis. In addition, we design a Scale-Aware Feature Synthesizer module to capture the dependencies between features at different scales, enhancing the model’s comprehension and utilization of multi-scale temporal information. Extensive experimental results on real-world datasets show that ARFNet outperforms state-of-the-art methods in time series forecasting.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 133044"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386722","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2026-05-01Epub Date: 2026-02-11DOI: 10.1016/j.neucom.2026.133015
Jianfeng Dong , Shengwei Tian , Long Yu , Hongfeng You , Qimeng Yang , Jinmiao Song , Xinjun Pei , Feng Shi , Kun Wu
Occluded person re-identification (ReID) faces an “information incompleteness paradox”: a single occluded view misses discriminative cues and yields ambiguous representations, while exploiting multi-view observations typically requires multi-branch inference with high computational cost. To address this dilemma, we propose Multi-View Consistency Distillation (MVCD), a framework formulated under the Learning Using Privileged Information (LUPI) paradigm. Specifically, we construct a training-only teacher that has privileged access to multi-view fragments and identity annotations (available only during training), and transfer this privileged knowledge to a standard single-view student through consistency distillation. The teacher contains three training-only mechanisms: (1) Saliency-Guided Feature Purification (SGFP) to suppress occlusion-induced noise with label guidance; (2) Cross-View Patch Alignment (CVPA) to exploit patch correspondences for spatial rectification across views; and (3) Reliability-Guided Aggregation (RGA) to produce a low-variance, reliable supervision target. Crucially, all auxiliary components are discarded after training, enabling the student to recover more complete representations from occluded inputs with zero extra inference-time cost. Extensive experiments on five benchmarks show consistent improvements over strong baselines. On Occluded-DukeMTMC, MVCD achieves 70.4% Rank-1 accuracy and runs at 25 ms per image, outperforming prior state-of-the-art methods.
{"title":"Learning from multi-view fragments: An adaptive consistency distillation framework for occluded person re-identification","authors":"Jianfeng Dong , Shengwei Tian , Long Yu , Hongfeng You , Qimeng Yang , Jinmiao Song , Xinjun Pei , Feng Shi , Kun Wu","doi":"10.1016/j.neucom.2026.133015","DOIUrl":"10.1016/j.neucom.2026.133015","url":null,"abstract":"<div><div>Occluded person re-identification (ReID) faces an “information incompleteness paradox”: a single occluded view misses discriminative cues and yields ambiguous representations, while exploiting multi-view observations typically requires multi-branch inference with high computational cost. To address this dilemma, we propose <strong>Multi-View Consistency Distillation (MVCD)</strong>, a framework formulated under the <strong>Learning Using Privileged Information (LUPI)</strong> paradigm. Specifically, we construct a <em>training-only</em> teacher that has privileged access to multi-view fragments and identity annotations (available only during training), and transfer this privileged knowledge to a standard single-view student through consistency distillation. The teacher contains three training-only mechanisms: (1) <strong>Saliency-Guided Feature Purification (SGFP)</strong> to suppress occlusion-induced noise with label guidance; (2) <strong>Cross-View Patch Alignment (CVPA)</strong> to exploit patch correspondences for spatial rectification across views; and (3) <strong>Reliability-Guided Aggregation (RGA)</strong> to produce a low-variance, reliable supervision target. Crucially, all auxiliary components are discarded after training, enabling the student to recover more complete representations from occluded inputs with <strong>zero extra inference-time cost</strong>. Extensive experiments on five benchmarks show consistent improvements over strong baselines. On Occluded-DukeMTMC, MVCD achieves 70.4% Rank-1 accuracy and runs at 25 ms per image, outperforming prior state-of-the-art methods.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 133015"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386736","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2026-05-01Epub Date: 2026-02-07DOI: 10.1016/j.neucom.2026.132977
Yuliang Cai , Mohammad Rostami
Transformer neural networks are increasingly replacing prior architectures across a wide range of applications in different data modalities. The increasing size and computational demands of fine-tuning large pre-trained transformer neural networks pose significant challenges for the widespread adoption of these models in applications that demand on-edge computing. To tackle this challenge, continual learning (CL) emerges as a solution by facilitating the transfer of knowledge across tasks that arrive sequentially for an autonomously learning agent. However, current CL methods mainly focus on learning tasks that are exclusively vision-based or language-based. We propose a transformer-based CL framework focusing on learning tasks that involve both vision and language, known as Vision-and-Language (VaL) tasks. In our framework, we benefit from the novel task-attention block and the introduced extra parameters to a base transformer to specialize the network for each task. As a result, we enable dynamic model expansion to learn several tasks in a sequence. We also use knowledge distillation and experience replay to benefit from relevant past experiences to learn the current task more efficiently. Our proposed method, Task Attentive Multimodal Continual Learning (TAM-CL), allows for the exchange of information between tasks while mitigating the problem of catastrophic forgetting. Notably, our approach is scalable, incurring minimal memory overhead. TAM-CL achieves 4.62% accuracy higher than the state-of-the-art (SOTA) accuracy on challenging multimodal tasks.2
{"title":"Dynamic transformer architecture for continual learning of multimodal tasks","authors":"Yuliang Cai , Mohammad Rostami","doi":"10.1016/j.neucom.2026.132977","DOIUrl":"10.1016/j.neucom.2026.132977","url":null,"abstract":"<div><div>Transformer neural networks are increasingly replacing prior architectures across a wide range of applications in different data modalities. The increasing size and computational demands of fine-tuning large pre-trained transformer neural networks pose significant challenges for the widespread adoption of these models in applications that demand on-edge computing. To tackle this challenge, continual learning (CL) emerges as a solution by facilitating the transfer of knowledge across tasks that arrive sequentially for an autonomously learning agent. However, current CL methods mainly focus on learning tasks that are exclusively vision-based or language-based. We propose a transformer-based CL framework focusing on learning tasks that involve both vision and language, known as Vision-and-Language (VaL) tasks. In our framework, we benefit from the novel task-attention block and the introduced extra parameters to a base transformer to specialize the network for each task. As a result, we enable dynamic model expansion to learn several tasks in a sequence. We also use knowledge distillation and experience replay to benefit from relevant past experiences to learn the current task more efficiently. Our proposed method, Task Attentive Multimodal Continual Learning (TAM-CL), allows for the exchange of information between tasks while mitigating the problem of catastrophic forgetting. Notably, our approach is scalable, incurring minimal memory overhead. TAM-CL achieves 4.62% accuracy higher than the state-of-the-art (SOTA) accuracy on challenging multimodal tasks.<span><span><sup>2</sup></span></span></div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 132977"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386377","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2026-05-01Epub Date: 2026-02-16DOI: 10.1016/j.neucom.2026.133069
Wentao Li, Xingwang Zhao, Zhiqiang Li, Shiyi Li, Jinghan Yang
Multi-view graph clustering (MGC) aims to integrate graph information from multiple perspectives to determine a unified node partitioning scheme, thereby accurately mining the potential association rules of data. In recent years, contrastive learning-based MGC methods have attracted extensive attention due to their excellent generalization ability. However, existing mainstream edge-based strategies (neighborhood contrastive learning) still have significant limitations: (1) The search range for positives is fixed to the first-order neighborhood, which is overly restrictive and makes it difficult to fully explore the latent similarity between nodes; (2) Excessive focus is placed on intra-view connection relationships while ignoring the mining of inter-view structural consistency, which tends to lead to the omission of positives with stable similarity to the anchor or the misclassification of low-similarity nodes as positives. To address the above limitations, this paper proposes a Multi-view Graph Contrastive Clustering via Consensus Constraint (MGC4). Specifically, the algorithm first leverages inter-view structural consistency to identify the consensus neighbors of anchors and their cross-view counterparts as high-quality positives. Second, it assigns adaptive contrastive weights according to the cross-view average graph distance between the positives and the anchor, distinguishing the importance of different positives. Extensive experiments on multiple real-world datasets demonstrate that MGC4 consistently outperforms current state-of-the-art baselines in both clustering accuracy and robustness.
{"title":"Multi-view graph contrastive clustering via consensus constraint","authors":"Wentao Li, Xingwang Zhao, Zhiqiang Li, Shiyi Li, Jinghan Yang","doi":"10.1016/j.neucom.2026.133069","DOIUrl":"10.1016/j.neucom.2026.133069","url":null,"abstract":"<div><div>Multi-view graph clustering (MGC) aims to integrate graph information from multiple perspectives to determine a unified node partitioning scheme, thereby accurately mining the potential association rules of data. In recent years, contrastive learning-based MGC methods have attracted extensive attention due to their excellent generalization ability. However, existing mainstream edge-based strategies (neighborhood contrastive learning) still have significant limitations: (1) The search range for positives is fixed to the first-order neighborhood, which is overly restrictive and makes it difficult to fully explore the latent similarity between nodes; (2) Excessive focus is placed on intra-view connection relationships while ignoring the mining of inter-view structural consistency, which tends to lead to the omission of positives with stable similarity to the anchor or the misclassification of low-similarity nodes as positives. To address the above limitations, this paper proposes a Multi-view Graph Contrastive Clustering via Consensus Constraint (MGC<sup>4</sup>). Specifically, the algorithm first leverages inter-view structural consistency to identify the consensus neighbors of anchors and their cross-view counterparts as high-quality positives. Second, it assigns adaptive contrastive weights according to the cross-view average graph distance between the positives and the anchor, distinguishing the importance of different positives. Extensive experiments on multiple real-world datasets demonstrate that MGC<sup>4</sup> consistently outperforms current state-of-the-art baselines in both clustering accuracy and robustness.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 133069"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2026-05-01Epub Date: 2026-02-12DOI: 10.1016/j.neucom.2026.132993
Bolin Song , Yiyang Wang , Changze Zhou
Self-supervised learning denoisers utilizing blind spot networks have garnered significant attention due to their capability to learn a denoiser that solely relies on single noisy images. However, existing approaches employ a uniform masking strategy, wherein all pixels are masked indiscriminately, resulting in the loss of intricate details. In this paper, we introduce a novel technique for generating a detail-guided mask that is not uniformly sampled across different regions, to mitigate the loss of intricate details and enhance overall denoising quality. Furthermore, we propose a novel framework named Self2Rolling to enhance the precision of detail position guidance and prioritize reinforcing crucial yet overlooked details that require preservation. The proposed framework can be viewed as an integration of a noise measurement model with self-supervised learning techniques, exhibiting continuous enhancements in denoising performance across iterations. Extensive experiments validate the superiority of our approach compared to state-of-the-art methods. The executable code will be released upon acceptance of the paper.
{"title":"Self2Rolling: Self-supervised denoising using detail-guided mask with updates","authors":"Bolin Song , Yiyang Wang , Changze Zhou","doi":"10.1016/j.neucom.2026.132993","DOIUrl":"10.1016/j.neucom.2026.132993","url":null,"abstract":"<div><div>Self-supervised learning denoisers utilizing blind spot networks have garnered significant attention due to their capability to learn a denoiser that solely relies on single noisy images. However, existing approaches employ a uniform masking strategy, wherein all pixels are masked indiscriminately, resulting in the loss of intricate details. In this paper, we introduce a novel technique for generating a detail-guided mask that is not uniformly sampled across different regions, to mitigate the loss of intricate details and enhance overall denoising quality. Furthermore, we propose a novel framework named Self2Rolling to enhance the precision of detail position guidance and prioritize reinforcing crucial yet overlooked details that require preservation. The proposed framework can be viewed as an integration of a noise measurement model with self-supervised learning techniques, exhibiting continuous enhancements in denoising performance across iterations. Extensive experiments validate the superiority of our approach compared to state-of-the-art methods. The executable code will be released upon acceptance of the paper.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 132993"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386365","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The stochastic gradient descent (SGD) algorithm has achieved remarkable success in training deep learning models. However, it has several limitations, including susceptibility to vanishing gradients, sensitivity to input data, and a lack of robust theoretical guarantees. In recent years, alternating minimization (AM) methods have emerged as a promising alternative for model training by employing gradient-free approaches to iteratively update model parameters. Despite their potential, these methods often exhibit slow convergence rates. To address this challenge, we propose a novel Triple-Inertial Accelerated Alternating Minimization (TIAM) framework for neural network training. The TIAM approach incorporates a triple-inertial acceleration strategy with a specialized approximation method, facilitating targeted acceleration of different terms in each sub-problem optimization. This integration improves the efficiency of convergence, achieving better performance with fewer iterations. Additionally, we provide a convergence analysis of the TIAM algorithm, including its global convergence properties and convergence rate. Extensive experiments validate the effectiveness of the TIAM method, demonstrating improvements in generalization capability and computational efficiency compared to existing approaches, particularly when applied to the rectified linear unit (ReLU) and its variants.
{"title":"A triple-inertial accelerated alternating optimization method for deep learning training","authors":"Chengcheng Yan, Jiawei Xu, Qingsong Wang, Zheng Peng","doi":"10.1016/j.neucom.2026.132997","DOIUrl":"10.1016/j.neucom.2026.132997","url":null,"abstract":"<div><div>The stochastic gradient descent (SGD) algorithm has achieved remarkable success in training deep learning models. However, it has several limitations, including susceptibility to vanishing gradients, sensitivity to input data, and a lack of robust theoretical guarantees. In recent years, alternating minimization (AM) methods have emerged as a promising alternative for model training by employing gradient-free approaches to iteratively update model parameters. Despite their potential, these methods often exhibit slow convergence rates. To address this challenge, we propose a novel Triple-Inertial Accelerated Alternating Minimization (TIAM) framework for neural network training. The TIAM approach incorporates a triple-inertial acceleration strategy with a specialized approximation method, facilitating targeted acceleration of different terms in each sub-problem optimization. This integration improves the efficiency of convergence, achieving better performance with fewer iterations. Additionally, we provide a convergence analysis of the TIAM algorithm, including its global convergence properties and convergence rate. Extensive experiments validate the effectiveness of the TIAM method, demonstrating improvements in generalization capability and computational efficiency compared to existing approaches, particularly when applied to the rectified linear unit (ReLU) and its variants.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"676 ","pages":"Article 132997"},"PeriodicalIF":6.5,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号