In multi-layer network community detection, the goal is to group nodes into distinct clusters based on their connection strengths. Currently, many existing methods do not fully leverage the relationships between layers, and observed multi-layer networks often contain noise that can significantly impact the accuracy of community detection. To address these challenges, a robust low-rank tensor constrained orthogonal symmetric non-negative matrix factorization method for multi-layer network community detection (RTOSNMF) is introduced. Specifically, noise is separated from raw adjacency matrices using linear separation, and a $l_{2,1}$ norm constraint is applied to achieve denoising. Clean adjacency matrices are then used to perform orthogonal symmetric non-negative matrix factorization, extracting latent representations of the multi-layer networks. Moreover, the nuclear norm is utilized to preserve the low-rank property of the adjacency tensor, aiding in the discovery of higher-order inter-layer relationships. An algorithm based on the Alternating Direction Method of Multipliers (ADMM) is designed to solve the RTOSNMF model. Extensive experiments conducted on eight datasets demonstrate superior performance of the proposed model compared with fifteen state-of-the-art methods.
{"title":"Robust Low-Rank Tensor Constrained Orthogonal Symmetric Non-Negative Matrix Factorization for Multi-Layer Networks Community Detection","authors":"Qianlong Zhou;Hangjun Che;Wei Guo;Xing He;Man-Fai Leung;Shiping Wen","doi":"10.1109/TETCI.2025.3572129","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3572129","url":null,"abstract":"In multi-layer network community detection, the goal is to group nodes into distinct clusters based on their connection strengths. Currently, many existing methods do not fully leverage the relationships between layers, and observed multi-layer networks often contain noise that can significantly impact the accuracy of community detection. To address these challenges, a robust low-rank tensor constrained orthogonal symmetric non-negative matrix factorization method for multi-layer network community detection (RTOSNMF) is introduced. Specifically, noise is separated from raw adjacency matrices using linear separation, and a <inline-formula><tex-math>$l_{2,1}$</tex-math></inline-formula> norm constraint is applied to achieve denoising. Clean adjacency matrices are then used to perform orthogonal symmetric non-negative matrix factorization, extracting latent representations of the multi-layer networks. Moreover, the nuclear norm is utilized to preserve the low-rank property of the adjacency tensor, aiding in the discovery of higher-order inter-layer relationships. An algorithm based on the Alternating Direction Method of Multipliers (ADMM) is designed to solve the RTOSNMF model. Extensive experiments conducted on eight datasets demonstrate superior performance of the proposed model compared with fifteen state-of-the-art methods.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 6","pages":"4145-4160"},"PeriodicalIF":5.3,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145584651","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}
This paper proposes a novel pooling-based VGG-Lite model in order to mitigate class imbalance issues in Chest X-Ray (CXR) datasets. Automatic Pneumonia detection from CXR images by deep learning model has emerged as a prominent and dynamic area of research, since the inception of the new Covid-19 variant in 2020. However, the standard Convolutional Neural Network (CNN) models encounter challenges associated with class imbalance, a prevalent issue found in many medical datasets. The innovations introduced in the proposed model architecture include: (I) A very lightweight CNN model, “VGG-Lite”, is proposed as a base model, inspired by VGG-16 and MobileNet-V2 architecture. (II) On top of this base model, we leverage an “Edge Enhanced Module (EEM)” through a parallel branch, consisting of a “negative image layer”, and a novel custom pooling layer “2Max-Min Pooling”. This 2Max-Min Pooling layer is entirely novel in this investigation, providing more attention to edge components within pneumonia CXR images. Thus, it works as an efficient spatial attention module (SAM). We have implemented the proposed framework on two separate CXR datasets. The first dataset is obtained from a readily available source on the internet, and the second dataset is a more challenging CXR dataset, assembled by our research team from three different sources. Experimental results reveal that our proposed framework has outperformed pre-trained CNN models, and three recent trend existing models “Vision Transformer”, “Pooling-based Vision Transformer (PiT)” and “PneuNet”, by substantial margins on both datasets. The proposed framework VGG-Lite with EEM, has achieved a macro average of 95% accuracy, 97.1% precision, 96.1% recall, and 96.6% F1 score on the “Pneumonia Imbalance CXR dataset”, without employing any pre-processing technique.
{"title":"Novel Pooling-Based VGG-Lite for Pneumonia and Covid-19 Detection From Imbalanced Chest X-Ray Datasets","authors":"Santanu Roy;Ashvath Suresh;Palak Sahu;Achintya Roy;Tulika Rudra Gupta","doi":"10.1109/TETCI.2025.3577509","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3577509","url":null,"abstract":"This paper proposes a novel pooling-based VGG-Lite model in order to mitigate class imbalance issues in Chest X-Ray (CXR) datasets. Automatic Pneumonia detection from CXR images by deep learning model has emerged as a prominent and dynamic area of research, since the inception of the new Covid-19 variant in 2020. However, the standard Convolutional Neural Network (CNN) models encounter challenges associated with class imbalance, a prevalent issue found in many medical datasets. The innovations introduced in the proposed model architecture include: (I) A very lightweight CNN model, “VGG-Lite”, is proposed as a base model, inspired by VGG-16 and MobileNet-V2 architecture. (II) On top of this base model, we leverage an “Edge Enhanced Module (EEM)” through a parallel branch, consisting of a “negative image layer”, and a novel custom pooling layer “2Max-Min Pooling”. This 2Max-Min Pooling layer is entirely novel in this investigation, providing more attention to edge components within pneumonia CXR images. Thus, it works as an efficient spatial attention module (SAM). We have implemented the proposed framework on two separate CXR datasets. The first dataset is obtained from a readily available source on the internet, and the second dataset is a more challenging CXR dataset, assembled by our research team from three different sources. Experimental results reveal that our proposed framework has outperformed pre-trained CNN models, and three recent trend existing models “Vision Transformer”, “Pooling-based Vision Transformer (PiT)” and “PneuNet”, by substantial margins on both datasets. The proposed framework VGG-Lite with EEM, has achieved a macro average of 95% accuracy, 97.1% precision, 96.1% recall, and 96.6% F1 score on the “Pneumonia Imbalance CXR dataset”, without employing any pre-processing technique.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 6","pages":"4231-4242"},"PeriodicalIF":5.3,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145584693","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 : 2025-06-05DOI: 10.1109/TETCI.2025.3573250
Yongxin Deng;Xihe Qiu;Xiaoyu Tan;Yaochu Jin
Reinforcement learning methods have been used to optimize long-term user engagement in recommendation systems. However, existing reinforcement learning-based recommendation systems do not fully exploit the relevance of individual user behavior across different platforms. One potential solution is to aggregate data from various platforms in a centralized location and use the aggregated data for training. However, this approach raises economic and legal concerns, including increased communication costs and potential threats to user privacy. To address these challenges, we propose FedSlate, a federated reinforcement learning recommendation algorithm that effectively utilizes information that is prohibited from being shared at a legal level. We employ the SlateQ algorithm to assist FedSlate in learning users' long-term behavior and evaluating the value of recommended content. We extend the existing application scope of recommendation systems from single-user single-platform to single-user multi-platform and address cross-platform learning challenges by introducing federated learning. We use RecSim to construct a simulation environment for evaluating FedSlate and compare its performance with state-of-the-art benchmark recommendation models. Experimental results demonstrate the superior effects of FedSlate over baseline methods in various environmental settings, and FedSlate facilitates the learning of recommendation strategies in scenarios where baseline methods are completely inapplicable.
{"title":"FedSlate: A Federated Deep Reinforcement Learning Recommender System","authors":"Yongxin Deng;Xihe Qiu;Xiaoyu Tan;Yaochu Jin","doi":"10.1109/TETCI.2025.3573250","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3573250","url":null,"abstract":"Reinforcement learning methods have been used to optimize long-term user engagement in recommendation systems. However, existing reinforcement learning-based recommendation systems do not fully exploit the relevance of individual user behavior across different platforms. One potential solution is to aggregate data from various platforms in a centralized location and use the aggregated data for training. However, this approach raises economic and legal concerns, including increased communication costs and potential threats to user privacy. To address these challenges, we propose <bold>FedSlate</b>, a federated reinforcement learning recommendation algorithm that effectively utilizes information that is prohibited from being shared at a legal level. We employ the SlateQ algorithm to assist FedSlate in learning users' long-term behavior and evaluating the value of recommended content. We extend the existing application scope of recommendation systems from single-user single-platform to single-user multi-platform and address cross-platform learning challenges by introducing federated learning. We use RecSim to construct a simulation environment for evaluating FedSlate and compare its performance with state-of-the-art benchmark recommendation models. Experimental results demonstrate the superior effects of FedSlate over baseline methods in various environmental settings, and FedSlate facilitates the learning of recommendation strategies in scenarios where baseline methods are completely inapplicable.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 6","pages":"4202-4216"},"PeriodicalIF":5.3,"publicationDate":"2025-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145584632","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 : 2025-06-05DOI: 10.1109/TETCI.2025.3573240
Haseena Rahmath P;Kuldeep Chaurasia;Anika
Graph Neural Networks (GNNs) are effective for learning on graph-structured data but often suffer from high inference costs, particularly in deeper architectures. Standard GNNs employ a single-exit design, processing all inputs through the entire network regardless of their complexity—resulting in unnecessary computation for simpler instances. This paper introduces E2G-Net, a multi-exit GNN architecture that inserts early-exit branches at intermediate layers to enable instance-adaptive inference. A Bayesian Optimization (BO)-based policy determines the optimal exit criterion and threshold at each branch, optimizing the trade-off between accuracy and efficiency. E2G-Net is evaluated using GCN and GAT backbones on ten node classification benchmarks spanning homophilic, heterophilic, and large-scale graphs. It achieves up to 3.7× inference speedup (Cornell) and over 45% FLOPs reduction (OGBN-Arxiv), while preserving and often improving classification accuracy across datasets. These results demonstrate E2G-Net's scalability and efficiency for real-world graph inference.
{"title":"E2G-Net: Enhancing Efficiency in Graph Neural Networks With Early-Exit Branches","authors":"Haseena Rahmath P;Kuldeep Chaurasia;Anika","doi":"10.1109/TETCI.2025.3573240","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3573240","url":null,"abstract":"Graph Neural Networks (GNNs) are effective for learning on graph-structured data but often suffer from high inference costs, particularly in deeper architectures. Standard GNNs employ a single-exit design, processing all inputs through the entire network regardless of their complexity—resulting in unnecessary computation for simpler instances. This paper introduces E2G-Net, a multi-exit GNN architecture that inserts early-exit branches at intermediate layers to enable instance-adaptive inference. A Bayesian Optimization (BO)-based policy determines the optimal exit criterion and threshold at each branch, optimizing the trade-off between accuracy and efficiency. E2G-Net is evaluated using GCN and GAT backbones on ten node classification benchmarks spanning homophilic, heterophilic, and large-scale graphs. It achieves up to 3.7× inference speedup (Cornell) and over 45% FLOPs reduction (OGBN-Arxiv), while preserving and often improving classification accuracy across datasets. These results demonstrate E2G-Net's scalability and efficiency for real-world graph inference.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 6","pages":"4191-4201"},"PeriodicalIF":5.3,"publicationDate":"2025-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145584670","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 : 2025-06-02DOI: 10.1109/TETCI.2025.3572125
Rong Wan;Feng-Feng Wei;Wei-Neng Chen
In social network analysis, the opinion maximization (OM) problem aims to locate several nodes as a seed set, which starts the information propagation and achieves the most positive opinion of a social network. Considering the situation in practice that decision makers prefer having alternatives to make the final decisions, the multimodal OM (mOM) problem is derived in this paper. In our work, the mOM problem is formulated first, with the goal to identify multiple heterogeneous well-performed seed sets for the primal OM problem. Secondly, a genetic algorithm with a novel niche technique, GA-ComFit, is developed to solve the proposed mOM problem. In GA-ComFit, potential seed sets are encoded as individuals. Built on fitness sharing, the community-based fitness sharing niching technique, ComFit, hierarchically clusters individuals into multiple niches based on the community feature of each individual. As a result, the proposed GA-ComFit generates multiple heterogeneous seed sets as the solution for the mOM problem. Furthermore, a series of experiments conducted on real-world social networks demonstrate that the proposed GA-ComFit generally offers a set of multiple excellent heterogeneous seed sets for the mOM problem. To the best of our knowledge, this is the first study of the OM problem from the multimodal optimization perspective.
{"title":"GA-ComFit: A Genetic Algorithm With Community Fitness Sharing Niching for Multimodal Opinion Maximization","authors":"Rong Wan;Feng-Feng Wei;Wei-Neng Chen","doi":"10.1109/TETCI.2025.3572125","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3572125","url":null,"abstract":"In social network analysis, the opinion maximization (OM) problem aims to locate several nodes as a seed set, which starts the information propagation and achieves the most positive opinion of a social network. Considering the situation in practice that decision makers prefer having alternatives to make the final decisions, the multimodal OM (mOM) problem is derived in this paper. In our work, the mOM problem is formulated first, with the goal to identify multiple heterogeneous well-performed seed sets for the primal OM problem. Secondly, a genetic algorithm with a novel niche technique, GA-ComFit, is developed to solve the proposed mOM problem. In GA-ComFit, potential seed sets are encoded as individuals. Built on fitness sharing, the community-based fitness sharing niching technique, ComFit, hierarchically clusters individuals into multiple niches based on the community feature of each individual. As a result, the proposed GA-ComFit generates multiple heterogeneous seed sets as the solution for the mOM problem. Furthermore, a series of experiments conducted on real-world social networks demonstrate that the proposed GA-ComFit generally offers a set of multiple excellent heterogeneous seed sets for the mOM problem. To the best of our knowledge, this is the first study of the OM problem from the multimodal optimization perspective.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 6","pages":"4161-4174"},"PeriodicalIF":5.3,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145584631","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 : 2025-06-02DOI: 10.1109/TETCI.2025.3572133
Min Huang;Zifeng Xie;Han Huang;Chang Zhang;Liuqi Zhao;Ziyan Feng
Multi-source unsupervised domain adaptation (MSUDA) is a technique that transfers knowledge from multiple labeled source domains to an unlabeled target domain. The challenge of MSUDA is to reduce the domain shift and effectively amalgamate knowledge from disparate source domains. To address this challenge, it is necessary to model the target domain as a weighted combination of the source domains at the category level. Therefore, we propose a prototype combination method for multi-source unsupervised domain adaptation, which establishes multiple domain alignment in a combinatorial manner. Our method is established on a set of semantic category prototypes, each of which is a representative category embedding. A prototype combination mechanism (i.e., a feature-fusion scheme) is designed to select which source class features should be aligned with the corresponding target class features. This method incorporates contrastive prototype adaptation (i.e., a category-wise alignment approach) to accommodate the label distributions of the target domain. Furthermore, a prototype combination regularization (i.e., a domain-wise alignment metric) is designed to reduce the distributional differences between the source category prototypes and the target samples of low-quality pseudo-labels. The experimental results on three benchmark datasets demonstrate that our prototype combination mechanism is capable of selecting and combining category-discriminative features across multiple source domains, while the prototype combination regularization can further reduce the domain shift.
{"title":"Prototype Combination for Multi-Source Unsupervised Domain Adaptation","authors":"Min Huang;Zifeng Xie;Han Huang;Chang Zhang;Liuqi Zhao;Ziyan Feng","doi":"10.1109/TETCI.2025.3572133","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3572133","url":null,"abstract":"Multi-source unsupervised domain adaptation (MSUDA) is a technique that transfers knowledge from multiple labeled source domains to an unlabeled target domain. The challenge of MSUDA is to reduce the domain shift and effectively amalgamate knowledge from disparate source domains. To address this challenge, it is necessary to model the target domain as a weighted combination of the source domains at the category level. Therefore, we propose a prototype combination method for multi-source unsupervised domain adaptation, which establishes multiple domain alignment in a combinatorial manner. Our method is established on a set of semantic category prototypes, each of which is a representative category embedding. A prototype combination mechanism (i.e., a feature-fusion scheme) is designed to select which source class features should be aligned with the corresponding target class features. This method incorporates contrastive prototype adaptation (i.e., a category-wise alignment approach) to accommodate the label distributions of the target domain. Furthermore, a prototype combination regularization (i.e., a domain-wise alignment metric) is designed to reduce the distributional differences between the source category prototypes and the target samples of low-quality pseudo-labels. The experimental results on three benchmark datasets demonstrate that our prototype combination mechanism is capable of selecting and combining category-discriminative features across multiple source domains, while the prototype combination regularization can further reduce the domain shift.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 6","pages":"4175-4190"},"PeriodicalIF":5.3,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145584634","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 : 2025-04-28DOI: 10.1109/TETCI.2025.3561629
Qi Zhao;Bai Yan;Taiwei Hu;Xianglong Chen;Jian Yang;Shi Cheng;Yuhui Shi
Metaheuristics are widely recognized gradient-free solvers to hard problems that do not meet the rigorous mathematical assumptions of conventional solvers. The automated design of metaheuristic algorithms provides an attractive path to relieve manual design effort and gain enhanced performance beyond human-made algorithms. However, the specific algorithm prototype and linear algorithm representation in the current automated design pipeline restrict the design within a fixed algorithm structure, which hinders discovering novelties and diversity across the metaheuristic family. To address this challenge, this paper proposes a general framework, AutoOpt, for automatically designing metaheuristic algorithms with diverse structures. AutoOpt contains three innovations: (i) A general algorithm prototype dedicated to covering the metaheuristic family as widely as possible. It promotes high-quality automated design on different problems by fully discovering potentials and novelties across the family. (ii) A directed acyclic graph algorithm representation to fit the proposed prototype. Its flexibility and evolvability enable discovering various algorithm structures in a single run of design, thus boosting the possibility of finding high-performance algorithms. (iii) A graph representation embedding method offering an alternative compact form of the graph to be manipulated, which ensures AutoOpt's generality. Experiments on numeral functions and real applications validate AutoOpt's efficiency and practicability.
{"title":"AutoOpt: A General Framework for Automatically Designing Metaheuristic Optimization Algorithms With Diverse Structures","authors":"Qi Zhao;Bai Yan;Taiwei Hu;Xianglong Chen;Jian Yang;Shi Cheng;Yuhui Shi","doi":"10.1109/TETCI.2025.3561629","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3561629","url":null,"abstract":"Metaheuristics are widely recognized gradient-free solvers to hard problems that do not meet the rigorous mathematical assumptions of conventional solvers. The automated design of metaheuristic algorithms provides an attractive path to relieve manual design effort and gain enhanced performance beyond human-made algorithms. However, the specific algorithm prototype and linear algorithm representation in the current automated design pipeline restrict the design within a fixed algorithm structure, which hinders discovering novelties and diversity across the metaheuristic family. To address this challenge, this paper proposes a general framework, AutoOpt, for automatically designing metaheuristic algorithms with diverse structures. AutoOpt contains three innovations: (i) A general algorithm prototype dedicated to covering the metaheuristic family as widely as possible. It promotes high-quality automated design on different problems by fully discovering potentials and novelties across the family. (ii) A directed acyclic graph algorithm representation to fit the proposed prototype. Its flexibility and evolvability enable discovering various algorithm structures in a single run of design, thus boosting the possibility of finding high-performance algorithms. (iii) A graph representation embedding method offering an alternative compact form of the graph to be manipulated, which ensures AutoOpt's generality. Experiments on numeral functions and real applications validate AutoOpt's efficiency and practicability.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 5","pages":"3690-3703"},"PeriodicalIF":5.3,"publicationDate":"2025-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145128526","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 : 2025-04-21DOI: 10.1109/TETCI.2025.3558447
Bowen Hu;Weiheng Yao;Sibo Qiao;Hieu Pham;Shuqiang Wang;Michael Kwok-Po Ng
In minimally-invasive brain surgeries with indirect and narrow operating environments, 3D brain reconstruction is crucial. However, as requirements of accuracy for some new minimally-invasive surgeries (such as brain-computer interface surgery) are higher and higher, the outputs of conventional 3D reconstruction, such as point cloud (PC), are facing the challenges that sample points are too sparse and the precision is insufficient. On the other hand, there is a scarcity of high-density point cloud datasets, which makes it challenging to train models for direct reconstruction of high-density brain point clouds. In this work, a novel model named stereoscopic-aware graph generative adversarial network (SG-GAN) with two stages is proposed to generate fine high-density PC conditioned on a single image. The Stage-I GAN sketches the primitive shape and basic structure of the organ based on the given image, yielding Stage-I point clouds. The Stage-II GAN takes the results from Stage-I and generates high-density point clouds with detailed features. The Stage-II GAN is capable of correcting defects and restoring the detailed features of the region of interest (ROI) through the up-sampling process. Furthermore, a parameter-free-attention-based free-transforming module is developed to learn the efficient features of input, while upholding a promising performance. Comparing with the existing methods, the SG-GAN model shows superior performance in terms of visual quality, objective measurements, and performance in classification, as demonstrated by comprehensive results measured by several evaluation metrics including PC-to-PC error and Chamfer distance.
{"title":"SG-GAN: Fine Stereoscopic-Aware Generation for 3D Brain Point Cloud Up-Sampling From a Single Image","authors":"Bowen Hu;Weiheng Yao;Sibo Qiao;Hieu Pham;Shuqiang Wang;Michael Kwok-Po Ng","doi":"10.1109/TETCI.2025.3558447","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3558447","url":null,"abstract":"In minimally-invasive brain surgeries with indirect and narrow operating environments, 3D brain reconstruction is crucial. However, as requirements of accuracy for some new minimally-invasive surgeries (such as brain-computer interface surgery) are higher and higher, the outputs of conventional 3D reconstruction, such as point cloud (PC), are facing the challenges that sample points are too sparse and the precision is insufficient. On the other hand, there is a scarcity of high-density point cloud datasets, which makes it challenging to train models for direct reconstruction of high-density brain point clouds. In this work, a novel model named stereoscopic-aware graph generative adversarial network (SG-GAN) with two stages is proposed to generate fine high-density PC conditioned on a single image. The Stage-I GAN sketches the primitive shape and basic structure of the organ based on the given image, yielding Stage-I point clouds. The Stage-II GAN takes the results from Stage-I and generates high-density point clouds with detailed features. The Stage-II GAN is capable of correcting defects and restoring the detailed features of the region of interest (ROI) through the up-sampling process. Furthermore, a parameter-free-attention-based free-transforming module is developed to learn the efficient features of input, while upholding a promising performance. Comparing with the existing methods, the SG-GAN model shows superior performance in terms of visual quality, objective measurements, and performance in classification, as demonstrated by comprehensive results measured by several evaluation metrics including PC-to-PC error and Chamfer distance.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 5","pages":"3677-3689"},"PeriodicalIF":5.3,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145128527","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 : 2025-04-14DOI: 10.1109/TETCI.2025.3535659
Hassan Gharoun;Navid Yazdanjue;Mohammad Sadegh Khorshidi;Fang Chen;Amir H. Gandomi
With the surge in data generation, both vertically (i.e., volume of data) and horizontally (i.e., dimensionality) the burden of the curse of dimensionality has become increasingly palpable. Feature selection, a key facet of dimensionality reduction techniques, has advanced considerably to address this challenge. One such advancement is the Boruta feature selection algorithm, which successfully discerns meaningful features by contrasting them to their permutated counterparts known as shadow features. Building on this, this paper introduces NeuroBoruta, that extends the traditional Boruta approach by integrating neural networks and calibration metrics to improve prediction accuracy and reduce model uncertainty. By augmenting shadow features with noise and utilizing neural network-based perturbation for importance evaluation, and further incorporating calibration metrics alongside accuracy this evolved version of the Boruta method is presented. Experimental results demonstrate that NeuroBoruta significantly enhances the predictive performance and reliability of classification models across various datasets, including medical imaging and standard UCI datasets. This study underscores the importance of considering both feature relevance and model uncertainty in the feature selection process, particularly in domains requiring high accuracy and reliability.
{"title":"Leveraging Neural Networks and Calibration Measures for Confident Feature Selection","authors":"Hassan Gharoun;Navid Yazdanjue;Mohammad Sadegh Khorshidi;Fang Chen;Amir H. Gandomi","doi":"10.1109/TETCI.2025.3535659","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3535659","url":null,"abstract":"With the surge in data generation, both vertically (i.e., volume of data) and horizontally (i.e., dimensionality) the burden of the curse of dimensionality has become increasingly palpable. Feature selection, a key facet of dimensionality reduction techniques, has advanced considerably to address this challenge. One such advancement is the Boruta feature selection algorithm, which successfully discerns meaningful features by contrasting them to their permutated counterparts known as shadow features. Building on this, this paper introduces NeuroBoruta, that extends the traditional Boruta approach by integrating neural networks and calibration metrics to improve prediction accuracy and reduce model uncertainty. By augmenting shadow features with noise and utilizing neural network-based perturbation for importance evaluation, and further incorporating calibration metrics alongside accuracy this evolved version of the Boruta method is presented. Experimental results demonstrate that NeuroBoruta significantly enhances the predictive performance and reliability of classification models across various datasets, including medical imaging and standard UCI datasets. This study underscores the importance of considering both feature relevance and model uncertainty in the feature selection process, particularly in domains requiring high accuracy and reliability.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 3","pages":"2179-2193"},"PeriodicalIF":5.3,"publicationDate":"2025-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10964323","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144148086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-04-14DOI: 10.1109/TETCI.2025.3555250
Jie Lin;Yuhao Ye;Shaobo Li;Hanlin Zhang;Peng Zhao
The sparse reward problem widely exists in multi-agent deep reinforcement learning, preventing agents from learning optimal actions and resulting in inefficient interactions with the environment. Many efforts have been made to design denser rewards and promote agent exploration. However, existing methods only focus on the breadth of action exploration, neglecting the rationality of action exploration in deep reinforcement learning, which leads to inefficient action exploration for agents. To address this issue, in this paper, we propose a novel curiosity-based action exploration method in incomplete information competition game environments, namely IGC, to improve both the breadth and rationality of action exploitation in multi-agent deep reinforcement learning for sparse-reward environments. Particularly, to enhance the capability of action exploration for agents, the distance reward is designed in our IGC method to increase the density of rewards in action exploration, thereby mitigating the sparse reward problem. In addition, by integrating the Intrinsic Curiosity Module (ICM) into DQN, we propose an enhanced ICM-DQN module, which enhances the breadth and rationality of subject action exploration for agents. By doing this, our IGC method can mitigate the randomness of the existing curiosity mechanism and increase the rationality of action exploration of agents, thereby enhancing the efficiency of action exploration. Finally, we evaluate the effectiveness of our IGC method on an incomplete information card game, namely Uno card game. The results demonstrate that our IGC method can achieve both better action exploration efficiency and greater winning-rate in comparison with existing methods.
{"title":"Improving Exploration in Deep Reinforcement Learning for Incomplete Information Competition Environments","authors":"Jie Lin;Yuhao Ye;Shaobo Li;Hanlin Zhang;Peng Zhao","doi":"10.1109/TETCI.2025.3555250","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3555250","url":null,"abstract":"The sparse reward problem widely exists in multi-agent deep reinforcement learning, preventing agents from learning optimal actions and resulting in inefficient interactions with the environment. Many efforts have been made to design denser rewards and promote agent exploration. However, existing methods only focus on the breadth of action exploration, neglecting the rationality of action exploration in deep reinforcement learning, which leads to inefficient action exploration for agents. To address this issue, in this paper, we propose a novel curiosity-based action exploration method in incomplete information competition game environments, namely IGC, to improve both the breadth and rationality of action exploitation in multi-agent deep reinforcement learning for sparse-reward environments. Particularly, to enhance the capability of action exploration for agents, the distance reward is designed in our IGC method to increase the density of rewards in action exploration, thereby mitigating the sparse reward problem. In addition, by integrating the Intrinsic Curiosity Module (ICM) into DQN, we propose an enhanced ICM-DQN module, which enhances the breadth and rationality of subject action exploration for agents. By doing this, our IGC method can mitigate the randomness of the existing curiosity mechanism and increase the rationality of action exploration of agents, thereby enhancing the efficiency of action exploration. Finally, we evaluate the effectiveness of our IGC method on an incomplete information card game, namely Uno card game. The results demonstrate that our IGC method can achieve both better action exploration efficiency and greater winning-rate in comparison with existing methods.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 5","pages":"3665-3676"},"PeriodicalIF":5.3,"publicationDate":"2025-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145128537","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：