IEEE Transactions on Emerging Topics in Computational Intelligence最新文献

{"title":"Event-Sampled Adaptive Fuzzy Control of MASS via Intermittent Position Data","authors":"Guibing Zhu;Yong Ma;Songlin Hu","doi":"10.1109/TETCI.2025.3526331","DOIUrl":"https://doi.org/10.1109/TETCI.2025.3526331","url":null,"abstract":"In this article, a novel event-sampled adaptive fuzzy control solution is developed for the tracking issue of maritime autonomous surface ships (MASS) under a cyber environment. In the control design, the network resources constraint, internal/external uncertainties and input saturations are involved. To save the issue of network resources, only the intermittent position data is transmitted for the control design of MASS, and an event-sampled adaptive fuzzy state observer (ESAFSO) that only depends on the intermittent position data is designed to recover the untransmitted velocity. The designed ESAFSO is independent of the design of control law, which can coordinate the network resource-saving and uncertainty compensation. In addition, to accommodate the effect resulting from the event sample, which makes the signal discontinuous required by control design, a new design method is developed by using the backstepping design framework with a multivariable dynamic surface control technique. Furthermore, a dynamic event triggering mechanism is established in the controller-actuator (C-A) channel, and a novel adaptive fuzzy output feedback control solution is developed. Under Lyapunov theory, it is indicated that, under the proposed event-sampled adaptive fuzzy control solution for the MASS only intermittent position data, all signals of the closed-loop control system are proved to be bounded. The effectiveness of the developed control solution is illustrated by simulations.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 4","pages":"3084-3096"},"PeriodicalIF":5.3,"publicationDate":"2025-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687695","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}

{"title":"Benchmarking Medical LLMs on Anesthesiology: A Comprehensive Dataset in Chinese","authors":"Bohao Zhou;Yibing Zhan;Zhonghai Wang;Yanhong Li;Chong Zhang;Baosheng Yu;Liang Ding;Hua Jin;Weifeng Liu;Xiongbin Wang;Dapeng Tao","doi":"10.1109/TETCI.2024.3502465","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3502465","url":null,"abstract":"With the recent success of large language models (LLMs), interest in developing them for medical domains has increased. However, due to the lack of benchmark datasets, evaluating the capabilities of medical LLMs remains challenging, particularly in highly specialized fields such as anesthesiology. To address this gap, we introduce a comprehensive anesthesiology benchmark dataset in Chinese, known as the Chinese Anesthesiology Benchmark (CAB). This benchmark facilitates the evaluation of medical LLMs for anesthesiology across three crucial dimensions: knowledge, application, and safety. Specifically, the CAB provides more than 8 k questions collected from examinations and books for knowledge-level evaluation; more than 2 k questions collected from online anesthesia consultations and hospitals for application-level evaluation; and 136 tests from seven anesthesia medical care scenarios for safety-level evaluation. With the proposed CAB dataset, we conducted a thorough evaluation of six medical LLMs, such as Bianque-2 and HuatuoGPT-13B, and eleven general LLMs, such as Qwen-7B-Chat and GPT-4. The evaluation results revealed that there are still clear gaps in the capacities of medical LLMs for anesthesiology compared with those of medical students in the field of anesthesia. We hope that the proposed CAB dataset can facilitate the development of medical LLMs for anesthesiology.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 4","pages":"3057-3071"},"PeriodicalIF":5.3,"publicationDate":"2025-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687762","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}

{"title":"Alzheimer's Disease Diagnosis Using Ensemble of Random Weighted Features and Fuzzy Least Square Twin Support Vector Machine","authors":"Rahul Sharma;Tripti Goel;M Tanveer;Mujahed Al-Dhaifallah","doi":"10.1109/TETCI.2024.3523714","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3523714","url":null,"abstract":"Alzheimer's disease (AD) is a devastating neurological condition affecting a significant portion of the world's aging population. Magnetic resonance imaging (MRI) has been widely adopted to visualize and analyze the structural atrophies and other brain deformities caused by AD. Due to the differences in brain anatomy, brain sub-regions such as grey matter (GM), white matter (WM), and cerebrospinal fluid (CSF) deteriorate. Research shows that changes in GM, WM, and CSF are among the earliest detectable AD markers, supporting their use in early diagnosis and monitoring disease progression. In this paper, GM, WM, and CSF have been extracted from the T1-weighted MRI scan acquired from the ADNI database. A fine-tuned DL model has been implemented for automated and all levels of feature extraction. For classification, the data points from the input space are explicitly translated into a randomized feature space using a neural network with randomly generated weights for the hidden layer. After feature projection, the extended features train classification models, where two non-parallel hyperplanes are optimized, with all associated parameters undergoing fuzzification to enhance model robustness. The proposed classifier, a randomized vectored fuzzy least square twin support vector machine, adeptly manages the challenges of uncertain, imbalanced, and nonlinear data commonly found in medical imaging. It integrates fuzzy membership functions to systematically address data uncertainty and employs a least squares formulation to optimize the model efficiently, ensuring high accuracy and scalability. The performance of the proposed model is tested and compared with popular state-of-the-art models, showing significant improvements in effectiveness.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 2","pages":"1281-1291"},"PeriodicalIF":5.3,"publicationDate":"2025-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143716411","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}

{"title":"Privacy-Preserving Learning Model Using Lightweight Encryption for Visual Sensing Industrial IoT Devices","authors":"B D Deebak;Seong Oun Hwang","doi":"10.1109/TETCI.2024.3523771","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3523771","url":null,"abstract":"Technological convergence in visual sensing with industrial IoT (VSI-IoT) can bring numerous advances to large-scale crowd management systems like visual crowdsensing. VSI-IoT has significant features, including sensing, computing, analyzing, and storing, to address the issues of bearing failures, such as unplanned outages, increased downtime, and reduced operational efficiency. By contrast, providing privacy to the IIoT environments is a challenging task. Thus, this paper presents a novel privacy-preserving learning (PPL) mechanism that senses the defect rate of bearing failures using lightweight model aggregation at edge computing systems to preserve the privacy features. This convergence model synthesizes shape features comprehensively to transform the feature vectors into predictive functions that examine the categorization models using a two-dimensional convolution neural network (2D-CNN). Using security analysis, we demonstrate that the proposed PPL can achieve better privacy protection and model accuracy to preserve the learning features without additional verifiability. Further, the examination results showed that the proposed 2D-CNN with BN and LN consumed less computation complexity to achieve better detection accuracy <inline-formula><tex-math>$(approx 87.91.9%$</tex-math></inline-formula> to <inline-formula><tex-math>$approx 99.98%)$</tex-math></inline-formula> and communication cost <inline-formula><tex-math>$ (approx 21.09MB$</tex-math></inline-formula> to <inline-formula><tex-math>$ 23.92MB)$</tex-math></inline-formula> over three bearing datasets (i.e., IMS-Rexnord, CWRU, and Paderborn) than other state-of-the-art approaches. Above all, the privacy preserving based AlexNet was implemented using CryptoNet and LoLa to show different sets of efficiencies such as processing time, privacy, and integrity checks to preserve system performance following time-sensitive application scenarios like supply-chain optimization.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 4","pages":"3039-3056"},"PeriodicalIF":5.3,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687763","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}

{"title":"A Novel Immune Algorithm for Multiparty Multiobjective Optimization","authors":"Kesheng Chen;Wenjian Luo;Qi Zhou;Yujiang liu;Peilan Xu;Yuhui Shi","doi":"10.1109/TETCI.2024.3515013","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3515013","url":null,"abstract":"Traditional multiobjective optimization problems (MOPs) are insufficiently equipped for scenarios involving multiple decision makers (DMs), which are prevalent in many practical applications. These scenarios are categorized as multiparty multiobjective optimization problems (MPMOPs). For MPMOPs, the goal is to find a solution set that is as close to the Pareto front of each DM as much as possible. This poses challenges for evolutionary algorithms in terms of searching and selecting. To better solve MPMOPs, this paper proposes a novel approach called the multiparty immune algorithm (MPIA). The MPIA incorporates an inter-party guided crossover strategy based on the individual's non-dominated sorting ranks from different DM perspectives and an adaptive activation strategy based on the proposed multiparty cover metric (MCM). These strategies enable MPIA to activate suitable individuals for the next operations, maintain population diversity from different DM perspectives, and enhance the algorithm's search capability. To evaluate the performance of MPIA, we compare it with ordinary multiobjective evolutionary algorithms (MOEAs) and state-of-the-art multiparty multiobjective optimization evolutionary algorithms (MPMOEAs) by solving synthetic multiparty multiobjective problems and real-world biparty multiobjective unmanned aerial vehicle path planning (BPUAV-PP) problems involving multiple DMs. Experimental results demonstrate that MPIA outperforms other algorithms.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 2","pages":"1238-1252"},"PeriodicalIF":5.3,"publicationDate":"2025-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143716517","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}

{"title":"Towards Hierarchical Temporal Excitation for Video Violence Recognition","authors":"Aihua Mao;Wanqing Wu;Wenwei Yan;Yuxiang Li;Haoxiang Wang","doi":"10.1109/TETCI.2024.3522201","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3522201","url":null,"abstract":"Video-based violence recognition has become a crucial research topic with the wide usage of surveillance cameras. However, recognizing violent behavior from video data is challenging because of the additional temporal dimension, the lack of a precise range of violent behavior, and the complex backgrounds that make recognizing the interaction between objects difficult. Previous works have ambiguous reasoning of temporal features and insufficient understanding of action relationships. To address these issues, we propose a hierarchical temporal excitation network, which is effective for learning deep object interactions in spatio-temporal information and utilizing the interaction to robustly identify violent behaviors even in complex scenarios. The model we proposed comprises of two modules for temporal excitation, namely the shift temporal adaptive module (STAM) and the sparse object interaction transformer module (SOI-Tr). STAM extracts coarse-grained temporal information by fusing the shift component with the temporal adaptive modeling component. Furthermore, considering that deep-layer temporal features are more conducive to network understanding, SOI-Tr is introduced to excite fine-grained temporal representation reasoning by critical object attention. We conduct extensive experiments on mainstream violence datasets and a new constructed multi-class violence (MCV) dataset. The results show that our method outperforms the state-of-the-art works and is superior in understanding the object interaction in violent behavior recognition.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 4","pages":"3025-3038"},"PeriodicalIF":5.3,"publicationDate":"2025-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687697","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}

{"title":"ESAI: Efficient Split Artificial Intelligence via Early Exiting Using Neural Architecture Search","authors":"Behnam Zeinali;Di Zhuang;J. Morris Chang","doi":"10.1109/TETCI.2024.3485677","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3485677","url":null,"abstract":"Deep neural networks have demonstrated superior performance in various computer vision tasks compared to traditional machine learning algorithms. However, deploying these models on resource-constrained mobile and IoT devices poses computational challenges. Many devices resort to cloud computing, where complex deep learning models analyze data on servers. This approach increases communication costs and hampers system efficiency in the absence of a network connection. In this paper, we introduce a novel framework for deploying deep neural networks on IoT devices. This framework leverages both cloud and on-device models by extracting meta-information from each sample's classification result. It assesses the classification's performance to determine whether sending the sample to the server is necessary. Extensive experiments on CIFAR10 and CINIC10 datasets reveal that only 45% of CIFAR10 and 60% of CINIC10 test data need to be transmitted to the server using this technique. The overall accuracy of the framework is 94% and 89%, respectively, enhancing the accuracy of both client and server models.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 1","pages":"961-971"},"PeriodicalIF":5.3,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143106792","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}

{"title":"Visual IoT Sensing Based on Robust Multilabel Discrete Signatures With Self-Topological Regularized Half Quadratic Lifting Functions","authors":"Bo-Wei Chen;Ying-Hsuan Wu","doi":"10.1109/TETCI.2024.3514999","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3514999","url":null,"abstract":"The Visual Internet of Things (VIoT) is a next-generation key component to smart cities as it provides more clues for backend analysis. However, the VIoT needs to preliminarily screen/tag data before dispatching them to edge nodes. Moreover, when incomplete data are present, e.g., continuously occluded images, mislabeling may occur, not to mention detecting multilabels accurately. Another challenge is that although pairwise similarities via graph regularization/embedding strengthen label estimation, it is hard to determine the weights/locality between pairwise samples. Besides, static weighting may create rigid topologies owing to fixed distances, which hardly reflect data locality and subsequently cause fitting biases. In view of such, this work presents a discrete signature generation method that can analyze and convert corrupted data into robust signatures, thereby maintaining VIoT recognition functionalities. In this study, half quadratic lifting functions are introduced to accommodate fitting biases caused by data corruption which cannot be effectively solved by <inline-formula><tex-math>$ell _{2}$</tex-math></inline-formula>, <inline-formula><tex-math>$ell _{1}$</tex-math></inline-formula>, and <inline-formula><tex-math>$ell _{2,1}$</tex-math></inline-formula> norms. To avoid rigid distances among pairwise samples and to enable flexible automatic weighting, self-topological graph embedding/regularization is proposed, where data locality can be dynamically adjusted/learned from data. Additionally, semirelaxation on discrete signatures is devised to avoid quadratic unconstrained binary optimization caused by self-topological graphs and to enable efficient discrete cyclic coordinate descent (DCCD) algorithms for discrete signature optimization. Finally, decorrelation and balancing constraints corresponding to the proposed method are developed. Experiments on open datasets showed that the proposed method yielded better mean average precision (mAP) than the baselines. The outcomes revealed improvements in mAP for distinct noise types, including continuous occlusion, missing values, and sample-specific outliers, with respective increases of <inline-formula><tex-math>$mathbf {10.97}boldsymbol{%}$</tex-math></inline-formula>, <inline-formula><tex-math>$mathbf {8.05}boldsymbol{%}$</tex-math></inline-formula>, and <inline-formula><tex-math>$mathbf {9.72}boldsymbol{%}$</tex-math></inline-formula>, individually. Such findings verified the effectiveness of the proposed idea.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 3","pages":"2531-2544"},"PeriodicalIF":5.3,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144148163","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}

{"title":"Optimal Navigation of an Automatic Guided Vehicle With Obstacle Constraints: A Broad Learning-Based Approach","authors":"Jialun Lai;Zongze Wu;Zhigang Ren;Qi Tan;Hanzhen Xiao","doi":"10.1109/TETCI.2024.3520481","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3520481","url":null,"abstract":"Trajectoryplanningis a critical component of realizing intelligence in autonomous unmanned systems. While learning-based control offers intriguing real-time control properties and generalization, it has shortcomings when it comes to long offline training requirements. To address this limitation, this paper introduces a real-time optimal navigation control approach based on Broad Learning (BL) for the trajectory planning problem of Automatic Guided Vehicle (AGV) in the presence of obstacle constraints. The proposed framework fully leverages the optimality achieved through traditional optimal control methods and the novel BL approach. To achieve this, the optimal control problem (OCP) is first constructed and then solved offline for the long-term trajectory planning problem for an AGV working within obstacle constraints. After that, an optimal state-control dataset is acquired with the task information embedding. Subsequently, the BL architecture is meticulously designed and trained offline using the dataset to acquire proficiency in the optimal state-control mapping, and a secondary improvement based on the spectral norm constraint is introduced into the original BL architecture. Consequently, this well-trained BL-based controller is proficiently employed to provide feedback control based on the AGV's current state. The numerical simulation section provides a comparative analysis of the network training time consumption for the BL method and the Deep Neural Network (DNN) method, and the impact of different feature representation methods on the BL-based controller is discussed. Additionally, a local re-planning framework for scenario alterations is proposed based on favourable performances. It further showcases the method's ability to mitigate the excessively long offline training times typically associated with learning-based approaches.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 4","pages":"3010-3024"},"PeriodicalIF":5.3,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687722","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}

{"title":"Robust Soft LVQ With LEML for SPD Matrices","authors":"Fengzhen Tang;Xiaocheng Zhang","doi":"10.1109/TETCI.2024.3515009","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3515009","url":null,"abstract":"Many learning scenarios involve non-Euclidean data. For instance, in electroencephalogram (EEG) or image classification, the input can be characterized by symmetric positive-definite (SPD) matrices. These matrices live on the curved Riemannian manifold instead of the flat Euclidean space. In such situations, classical Euclidean learning methods may fail due to the non-Euclidean nature of the data. This article proposes to generalize robust soft learning vector quantization (RSLVQ) targeted for Euclidean data to cope with such data in the log-Euclidean framework. Log-Euclidean metric learning (LEML) is incorporated into the RSLVQ framework, jointly learning the prototypes on the manifold and the distance metric tensor of the tangent map that projects the original tangent space to a more discriminative one. Two methods are subsequently proposed to learn the distance metric tensor. It is firstly confined to have full rank and treated as an SPD matrix, which is learned using the log-Euclidean framework. Then, this constraint is removed letting it become a symmetric positive semidefinite (SPSD) matrix, which is learned using the quotient geometry. In addition, we propose to adapt the variance parameter introduced in the probabilistic modelling during the training course of the classifier by the minimization of the negative log likelihood function with respect to this parameter. Experiments on multiple data sets with different properties show the proposed methods have good classification performances and low computational complexities.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"9 4","pages":"2995-3009"},"PeriodicalIF":5.3,"publicationDate":"2024-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10816565","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687730","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}