{"title":"Kolmogorov-Arnold Vision Transformer for Image Reconstruction in Lung Electrical Impedance Tomography","authors":"Ibrar Amin;Shuaikai Shi;Hasan AlMarzouqi;Zeyar Aung;Waqar Ahmed;Panos Liatsis","doi":"10.1109/OJCS.2025.3559390","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3559390","url":null,"abstract":"Electrical impedance tomography is a non-invasive and non-ionizing imaging technique, which can provide real-time monitoring of the internal structures and function of the human body, and has been particularly popular in lung monitoring. However, the associated inverse problem is ill-posed, leading to suboptimal image quality with low spatial resolution, which hinders its practical use in the clinical settings. To achieve reliable image reconstruction, this work proposes a novel deep learning approach, applied to lung monitoring. The proposed model is a hybrid of the vision transformer and the recently introduced Kolmogorov Arnold Network (KAN). The fully connected layers in the transformer are replaced with KAN layers, which enhances its ability to learn the complex relationship between the voltage measurements and the conductivity distribution within the lungs. In comparison with the use of convolutional models and Vision Transformer, the proposed method achieves outstanding performance with a mean squared error of 0.0045, structural similarity index of 0.96, relative error of 0.11, and correlation coefficient of 0.98.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"519-530"},"PeriodicalIF":0.0,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10960369","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143896479","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Robust Joint Active and Passive Beamforming for Reconfigurable Intelligent Surface Assisted Full-Duplex Transmissions Under Imperfect Channels","authors":"Li-Hsiang Shen;Chia-Jou Ku;Kai-Ten Feng","doi":"10.1109/OJCS.2025.3556710","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3556710","url":null,"abstract":"The sixth-generation (6G) wireless technology recognizes the potential of reconfigurable intelligent surfaces (RIS) as an effective technique for intelligently manipulating channel paths through reflection to serve desired users. Full-duplex (FD) systems, enabling simultaneous transmission and reception from a base station (BS), offer the theoretical advantage of doubled spectrum efficiency. However, the presence of strong self-interference (SI) in FD systems significantly degrades performance, which can be mitigated by leveraging the capabilities of RIS. Moreover, accurately obtaining channel state information (CSI) from RIS poses a critical challenge. Our objective is to maximize downlink (DL) user data rates while ensuring quality-of-service (QoS) for uplink (UL) users under imperfect CSI from reflected channels. To address this, we propose a robust active BS and passive RIS beamforming (RAPB) scheme for RIS-FD, accounting for both SI and imperfect CSI. RAPB incorporates distributionally robust design, conditional value-at-risk (CVaR), and penalty convex-concave programming (PCCP) techniques. Simulation results demonstrate the UL/DL rate improvement are achieved by considering different levels of imperfect CSI. The proposed RAPB schemes validate their effectiveness across different RIS deployments and RIS/BS configurations. Benefited from robust beamforming, RAPB outperforms the existing methods in terms of non-robustness, deployment without RIS, conventional approximation, and half-duplex systems.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"502-518"},"PeriodicalIF":0.0,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10946837","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143892521","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Unbounded Depth ElGamal-Based Asymmetric Updatable Encryption Technique","authors":"Mostefa Kara;Ammar Boukrara;Mohammad Hammoudeh;Muhamad Felemban;Samir Guediri","doi":"10.1109/OJCS.2025.3551877","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3551877","url":null,"abstract":"This article introduces an ElGamal-based asymmetric updatable encryption scheme, tailored to address the challenges of secure key rotation in cryptographic systems. The proposed solution enables ciphertexts encrypted under an old key to be efficiently and securely updated to a new key without decryption, ensuring data confidentiality and integrity. By leveraging ElGamal's inherent mathematical properties, the scheme provides unbounded depth for key updates, asymmetric encryption capabilities, and independence from specific ciphertext structures. Lightweight pseudorandom generators (PRGs) are used to facilitate secure and efficient management of the random values required for encryption and re-encryption processes. The proposed approach demonstrates robust forward and backward security, ensuring resilience against information leakage even in the event of key compromise. Comprehensive performance evaluations highlight its efficiency, with minimal computational and communication overhead, making it suitable for large-scale systems and resource-constrained environments. Comparative analysis further confirms its superiority over existing techniques in encryption speed, ciphertext update time, and scalability. This work provides a practical and secure framework for managing frequent key updates in diverse applications, including cloud storage, the Internet of Things, and secure communication networks.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"491-501"},"PeriodicalIF":0.0,"publicationDate":"2025-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10938675","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143883489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"An Enhanced Deep Learning Approach to Potential Purchaser Prediction: AutoGluon Ensembles for Cross-Industry Profit Maximization","authors":"Hashibul Ahsan Shoaib;Md Anisur Rahman;Jannatul Maua;Ashifur Rahman;M. F. Mridha;Pankoo Kim;Jungpil Shin","doi":"10.1109/OJCS.2025.3552376","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3552376","url":null,"abstract":"Accurately identifying potential purchasers is critical for maximizing profitability in highly competitive markets, spanning industries from finance and insurance to telecommunications. This article presents an enhanced deep learning approach for potential purchaser prediction, leveraging an AutoGluon ensemble framework to optimize accuracy and profitability across diverse datasets, including time deposits, health insurance, 5G packages, and credit cards. The proposed AutoGluon-based ensemble integrates neural networks with boosted trees, stacking, and bagging to maximize the Expected Maximum Profit Criterion (EMPC) and deliver consistent predictive performance across datasets. Our model demonstrates superior performance in terms of Area Under the Curve (AUC), EMPC, and top decile lift (TDL) relative to benchmark classifiers. Specifically, for the credit card dataset, the model achieved an AUC of 0.8856, an EMPC of 13.8453, and a TDL of 3.80, marking significant improvements over prior results. Bayesian A/B testing, based on 40 EMPC ranks, further confirms the robustness of our model, with a 98.5% probability of being the best-performing model across datasets. The AutoGluon ensemble consistently outperforms traditional ensemble models, achieving an average rank-adjusted p-value below 0.015 in the Holm post-hoc test, validating its statistical significance. This study underscores the efficacy of deep learning ensembles in cross-industry potential purchaser prediction, providing a scalable, profit-driven approach for enhanced marketing and customer acquisition strategies.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"468-479"},"PeriodicalIF":0.0,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10930799","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817949","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Editorial: Gratitude, Reflection, and Celebration: My Tenure as EiC Comes to a Close","authors":"Song Guo","doi":"10.1109/OJCS.2025.3525990","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3525990","url":null,"abstract":"","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"1-1"},"PeriodicalIF":0.0,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10924579","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143602012","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Reference-Free 3D WiFi AP Localization by Outdoor-to-Indoor Bridging","authors":"Tatsuya Amano;Hirozumi Yamaguchi;Teruo Higashino","doi":"10.1109/OJCS.2025.3566774","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3566774","url":null,"abstract":"WiFi access point (AP) localization is essential for wireless infrastructure management and location-based services. While deep learning approaches have shown promising accuracy improvements, they require extensive training data with precise coordinates, making large-scale deployment impractical. Traditional localization techniques also rely heavily on indoor reference points (RPs), resulting in costly and labor-intensive deployments. We present WiSight, a novel framework that eliminates the need for indoor RPs by leveraging GPS-tagged outdoor RSS measurements and 3D building geometry, anchoring the indoor reference frame to the global coordinate system using GPS-tagged exterior points. WiSight first identifies virtual anchor positions on building exteriors through outdoor signal propagation modeling, then reconstructs indoor AP configurations using unlabeled RSS measurement pairs and multidimensional scaling. Extensive evaluation across multiple buildings demonstrates that WiSight achieves an average 3D AP localization error of 7.1 m (median: 6.8 m), reducing error by 59% compared to an opportunistic GPS-based approach. In office environments, WiSight attains 9.6 m error (median: 8.5 m)—22% lower than the state-of-the-art deep learning-based method, while achieving 82% floor-level accuracy without requiring any indoor RPs or training data.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"688-700"},"PeriodicalIF":0.0,"publicationDate":"2025-03-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10982448","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144125492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Correlation-Based Knowledge Distillation in Exemplar-Free Class-Incremental Learning","authors":"Zijian Gao;Bo Liu;Kele Xu;Xinjun Mao;Huaimin Wang","doi":"10.1109/OJCS.2025.3546754","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3546754","url":null,"abstract":"Class-incremental learning (CIL) aims to learn a family of classes incrementally with data available in order rather than training all data at once. One main drawback of CIL is that standard deep neural networks suffer from catastrophic forgetting (CF), especially when the model only has access to data from the current incremental step. Knowledge Distillation (KD) is a widely used technique that utilizes old models as the teacher model to alleviate CF. However, based on a case study, our investigation reveals that the vanilla KD is insufficient with a strict point-to-point restriction. Instead, a relaxed match between the teacher and student improves distillation performance and model stability. In this article, we propose a simple yet effective method to mitigate CF without any additional training costs or requiring any exemplars. Specifically, we apply the linear correlation between the features of the teacher and student to measure the distillation loss rather than vanilla point-to-point loss, which significantly improves the model stability. Then, we utilize label augmentation to improve feature generalization and save prototypes to alleviate classification bias further. The proposed method significantly outperforms state-of-the-art methods in the various settings of benchmarks, including CIFAR-100 and Tiny-ImageNet, demonstrating its effectiveness and robustness.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"449-459"},"PeriodicalIF":0.0,"publicationDate":"2025-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10908063","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143698225","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Nonparametric Bootstrap Likelihood Estimation to Investigate the Chance Set-Up on Clustering Results","authors":"Ammar Elnour;Wencheng Yang;Yan Li","doi":"10.1109/OJCS.2025.3545261","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3545261","url":null,"abstract":"Clustering algorithms are widely used in the knowledge discovery domain, but concerns and questions about the validity of the results must be considered. The datasets commonly used for clustering tasks are often large and scale-free, making conventional statistical techniques inadequate for analyzing result uncertainty. This issue applies to most outcomes obtained from other knowledge discovery techniques, such as machine learning and statistical learning. Traditional statistical methods assume data follows standard distributions, whereas resampling and bootstrapping methods offer more accurate and reliable alternatives. This article introduces a method that employs bootstrap likelihood estimation to infer the uncertainty of generated clustering structures. We first calculated the clustering error in the original dataset and then utilized the proposed method to estimate its nonparametric bootstrapped likelihood. By comparing these two values, we can establish a nonparametric significance testing framework that directly determines the validity of the result. To evaluate the effectiveness of our method, we conducted experiments using synthetic and real datasets. The results demonstrate that our method can successfully validate clustering results.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"438-448"},"PeriodicalIF":0.0,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10902121","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143698199","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"FraudGNN-RL: A Graph Neural Network With Reinforcement Learning for Adaptive Financial Fraud Detection","authors":"Yiwen Cui;Xu Han;Jiaying Chen;Xinguang Zhang;Jingyun Yang;Xuguang Zhang","doi":"10.1109/OJCS.2025.3543450","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3543450","url":null,"abstract":"As financial systems become increasingly complex and interconnected, traditional fraud detection methods struggle to keep pace with sophisticated fraudulent activities. This article introduces FraudGNN-RL, an innovative framework that combines Graph Neural Networks (GNNs) with Reinforcement Learning (RL) for adaptive and context-aware financial fraud detection. Our approach models financial transactions as a dynamic graph, where entities (e.g., users, merchants) are nodes and transactions form edges. We propose a novel GNN architecture, Temporal-Spatial-Semantic Graph Convolution (TSSGC), which simultaneously captures temporal patterns, spatial relationships, and semantic information in transaction data. The RL component, implemented as a Deep Q-Network (DQN), dynamically adjusts the fraud detection threshold and feature importance, allowing the model to adapt to evolving fraud patterns and minimize detection costs. We further introduce a Federated Learning scheme to enable collaborative model training across multiple financial institutions while preserving data privacy. Extensive experiments on a large-scale, real-world financial dataset demonstrate that FraudGNN-RL outperforms state-of-the-art baselines, achieving a 97.3% F1-score and reducing false positives by 31% compared to the best-performing baseline. Our framework also shows remarkable resilience to concept drift and adversarial attacks, maintaining high performance over extended periods. These results suggest that FraudGNN-RL offers a robust, adaptive, and privacy-preserving solution for financial fraud detection in the era of Big Data and interconnected financial ecosystems.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"426-437"},"PeriodicalIF":0.0,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892045","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143792930","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Threats, Attacks, and Defenses in Machine Unlearning: A Survey","authors":"Ziyao Liu;Huanyi Ye;Chen Chen;Yongsen Zheng;Kwok-Yan Lam","doi":"10.1109/OJCS.2025.3543483","DOIUrl":"https://doi.org/10.1109/OJCS.2025.3543483","url":null,"abstract":"Machine Unlearning (MU) has recently gained considerable attention due to its potential to achieve Safe AI by removing the influence of specific data from trained Machine Learning (ML) models. This process, known as knowledge removal, addresses AI governance concerns of training data such as quality, sensitivity, copyright restrictions, obsolescence, and compliance with privacy regulations. Efforts have been made to design efficient unlearning approaches, with MU services being examined for integration with existing machine learning as a service (MLaaS), allowing users to submit requests to remove specific data from the training corpus. However, recent research highlights vulnerabilities in machine unlearning systems that can lead to significant security and privacy concerns. Moreover, extensive research indicates that unlearning methods and prevalent attacks fulfill diverse roles within MU systems. This underscores the intricate relationship and complex interplay among these mechanisms in maintaining system functionality and safety. This survey aims to fill the gap between the extensive number of studies on threats, attacks, and defenses in machine unlearning and the absence of a comprehensive review that categorizes their taxonomy, methods, and solutions, thus offering valuable insights for future research directions and practical implementations.","PeriodicalId":13205,"journal":{"name":"IEEE Open Journal of the Computer Society","volume":"6 ","pages":"413-425"},"PeriodicalIF":0.0,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892039","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143645299","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}