计算机科学最新文献

{"title":"Physics-informed partitioned coupled neural operator for complex networks","authors":"Weidong Wu ,&nbsp;Yong Zhang ,&nbsp;Lili Hao ,&nbsp;Yang Chen ,&nbsp;Xiaoyan Sun ,&nbsp;Dunwei Gong","doi":"10.1016/j.engappai.2025.111567","DOIUrl":"10.1016/j.engappai.2025.111567","url":null,"abstract":"<div><div>Physics-Informed Neural Operators provide efficient, high-fidelity simulations for systems governed by partial differential equations. However, most existing studies focus only on multi-scale, multi-physics systems within a single spatial region, neglecting the case with multiple interconnected sub-regions, such as gas and thermal network systems. To address this, this paper proposes a Physics-Informed Partitioned Coupled Neural Operator to enhance the simulation performance of such networks. Compared to the existing Fourier Neural Operator, this method designs a joint convolution operator within the Fourier layers, enabling global integration capturing all sub-regions. Additionally, grid alignment layers are introduced outside the Fourier layers to help the joint convolution operator accurately learn the coupling relationship between sub-regions in the frequency domain. Experiments on gas, petroleum and transportation networks demonstrate that the proposed operator not only accurately simulates these complex networks but also shows good generalization and low model complexity.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"159 ","pages":"Article 111567"},"PeriodicalIF":7.5,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144534295","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}

{"title":"Coordinate plane based authentication method for detecting clone node in wireless sensor networks","authors":"Weinan Liu ,&nbsp;Jiawen Shi ,&nbsp;Hong Wang ,&nbsp;Tingting Chen ,&nbsp;Zhaoyang Han ,&nbsp;Qingqing Li","doi":"10.1016/j.jisa.2025.104148","DOIUrl":"10.1016/j.jisa.2025.104148","url":null,"abstract":"<div><div>Nowadays, wireless sensor networks (WSNs) have become a very promising technology for automatic data collection in many applications. Due to the feature of limited resource, WSNs are more vulnerable to certain attacks, such as node clone attacks. An adversary can clone a valid member sensor node and place the new clone node within the group to collect information in the group. The clone node has the same information as the cloned one, and can act as if it were the cloned one to obtain the group key, leading to leakage of group communication data. The current solutions have drawbacks; for instance, schemes based on IDS require additional component support. In this paper, a novel authentication scheme is proposed to address node clone attacks, utilizing a coordinate plane instead of geographical locations. This scheme also possesses additional functionalities, effectively managing node additions and revocations while incorporating collusion attack detection. Through theoretical analysis, the detection rate of our scheme is approximately 99.5%. Experimental simulations demonstrate that the practical detection rate of our scheme is 98.4%, which is lower than the theoretical maximum rate but is higher than that of many recent works and does not rely on additional mechanisms such as trust or hierarchical structures. Furthermore, through multiple rounds of detection, the overall detection rate can be further improved, and collusion attacks can be effectively identified.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"93 ","pages":"Article 104148"},"PeriodicalIF":3.8,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144534635","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}

{"title":"Lyapunov-based adaptive deep system identification for approximate dynamic programming","authors":"Wanjiku A. Makumi ,&nbsp;Omkar Sudhir Patil ,&nbsp;Warren E. Dixon","doi":"10.1016/j.automatica.2025.112462","DOIUrl":"10.1016/j.automatica.2025.112462","url":null,"abstract":"<div><div>Recent developments in approximate dynamic programming (ADP) use deep neural network (DNN)-based system identifiers to solve the infinite horizon state regulation problem; however, the DNN weights do not continually adjust for all layers. In this paper, ADP is performed using a Lyapunov-based DNN (Lb-DNN) adaptive identifier that involves online weight updates. Provided the Jacobian of the Lb-DNN satisfies the persistence of excitation condition, the Lb-DNN weights exponentially converge to a residual approximation error, and the corresponding control policy converges to a neighborhood of the optimal policy. Simulation results show that the Lb-DNN yields 49.85% improved root mean squared (RMS) function approximation error in comparison to a baseline ADP DNN result and faster convergence of the RMS regulation error, RMS controller error, and RMS function approximation error.</div></div>","PeriodicalId":55413,"journal":{"name":"Automatica","volume":"180 ","pages":"Article 112462"},"PeriodicalIF":4.8,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144534961","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}

{"title":"Double stage network inference system using k means clustering and fuzzy cognitive maps for cardiovascular disease diagnosis","authors":"Stephen Mariadoss,&nbsp;Felix Augustin","doi":"10.1016/j.engappai.2025.111540","DOIUrl":"10.1016/j.engappai.2025.111540","url":null,"abstract":"<div><div>The diagnosis of cardiovascular disease (CVD) is crucial as it stands as a prominent cause of mortality and its risk factors are often modifiable. Early diagnosis is an essential component of overall CVD health management and prevention strategies. When diagnosing CVD, numerous uncertainties arise, including variations in symptoms among individuals and symptom overlap with other diseases. To deal such uncertainties in the diagnosis of CVD, a fuzzy logic based diagnostic system is required. While designing fuzzy inference system (FIS), optimizing the rule is challenging task. When the rules are optimized in the system, the time complexity can be minimized and performance of the system may also be effective. Since optimized inference system will be more efficient, accurate, interpretable and easier to maintain, the objective of the study is to design a novel hybrid double stage network inference system using <span><math><mi>k</mi></math></span> means clustering, fuzzy cognitive maps (FCM) and Mamdani fuzzy inference system (MFIS) for diagnosing CVD. Initially, risk factors of CVD are categorized into modifiable and non-modifiable factors through <span><math><mi>k</mi></math></span> means clustering. Then, most influencing risk factors are identified from the modifiable risk factors using FCMs. The rules are created by implementing double stage network by incorporating sub factors of most influencing factors and biological factors. Then, the obtained rules from the double stage network are integrated into a MFIS to determine the level of CVD. The system’s effectiveness is assessed using a real-time clinical dataset comprising 1250 CVD risks, employing performance metrics, receiver operating curve (ROC) analysis and statistical evaluations. The proposed system demonstrates an exceptional performance, achieving a 99.23% accuracy, 98.99% sensitivity, 95.76% specificity and 99.46% precision in detecting CVD. This piece of work suggested that the proposed technique serves as a valuable tool for diagnosing CVD risks.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"159 ","pages":"Article 111540"},"PeriodicalIF":7.5,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144534453","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}

{"title":"A comprehensive review of cybersecurity vulnerabilities, threats, and solutions for the Internet of Things at the network-cum-application layer","authors":"Abdul Razaque ,&nbsp;Salim Hariri ,&nbsp;Abrar M. Alajlan ,&nbsp;Joon Yoo","doi":"10.1016/j.cosrev.2025.100789","DOIUrl":"10.1016/j.cosrev.2025.100789","url":null,"abstract":"<div><div>The proliferation of smart homes, smart logistics, and other technologies has expedited the expansion of Internet-of-Things (IoT) devices. This expansion has heightened the complexity of associated security challenges. Despite extensive research on IoT security, several studies fail to provide a comprehensive examination of both the network and application layers. This is particularly applicable to real-time and mission-critical settings. This review addresses that deficiency by offering a systematic review of IoT across five tiers. It concentrates on the application layer, categorizing it into three domains: real-time control systems, scientific decision-making systems, and query/scan search systems. The study examines vulnerabilities, attack vectors, and security measures in real-time control and query/scan systems. It examines how emerging technologies such as artificial intelligence (AI), Software Defined Networking (SDN), and fog/edge computing can enhance security via improved context awareness and access management. The study ultimately presents recommendations and suggests enhancements to foster trust, scalability, and enhanced security in contemporary IoT systems.</div></div>","PeriodicalId":48633,"journal":{"name":"Computer Science Review","volume":"58 ","pages":"Article 100789"},"PeriodicalIF":13.3,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144536154","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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":"SDVD: Self-supervised dual-view modeling of user and cascade dynamics for information diffusion prediction","authors":"Haoyu Xiong ,&nbsp;Jiaxing Shang ,&nbsp;Fei Hao ,&nbsp;Dajiang Liu ,&nbsp;Geyong Min","doi":"10.1016/j.knosys.2025.114005","DOIUrl":"10.1016/j.knosys.2025.114005","url":null,"abstract":"<div><div>Information diffusion prediction aims to estimate the likelihood of a user participating in a spreading message by leveraging social relationships and historical diffusion patterns. However, existing approaches often overlook a crucial factor: users’ participation behaviors are influenced by diverse and evolving motivations. Moreover, treating historical data as a whole may introduce noise from outdated information — especially as new users join — highlighting the dynamic nature of diffusion cascades. In addition, many current methods lack explicit supervision signals to effectively model these dynamics. To address these limitations, we propose SDVD, a novel framework for <strong>S</strong>elf-supervised <strong>D</strong>ual-<strong>V</strong>iew modeling of user and cascade <strong>D</strong>ynamics for information diffusion prediction. SDVD begins by constructing two auxiliary graphs from historical data: an adjacency dependency graph to capture temporal dependencies and a hypergraph to model group interactions. These structures explicitly model cascade dynamics and enhance user–cascade interaction understanding. We leverage graph neural networks and hypergraph neural networks to extract structural features from the graphs and introduce a user-aware fusion mechanism that integrates multisource information while reducing redundancy and noise. Furthermore, we design a self-supervised dual-view dynamic modeling module to learn temporal variations in diffusion patterns from both user and cascade perspectives. A cross-attention mechanism then combines static and dynamic representations, capturing contextual information within the cascade sequence. Experiments on four real-world datasets — with consistent preprocessing and data splitting — show that SDVD achieves statistically significant improvements (<span><math><mrow><mi>p</mi><mo>&lt;</mo><mn>0</mn><mo>.</mo><mn>05</mn></mrow></math></span>), with up to a 6.63% increase in MAP@10.</div></div>","PeriodicalId":49939,"journal":{"name":"Knowledge-Based Systems","volume":"326 ","pages":"Article 114005"},"PeriodicalIF":7.2,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144549715","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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":"Design and Characterisation of All-Dielectric Metasurface Reflector for mmWave Antennas","authors":"Muhammad Uzair,&nbsp;Guillaume François,&nbsp;Dirk Heberling,&nbsp;Suramate Chalermwisutkul","doi":"10.1049/mia2.70039","DOIUrl":"https://doi.org/10.1049/mia2.70039","url":null,"abstract":"<p>This study presents the design, characterisation and prototyping of a novel all-dielectric artificial magnetic conductor (AMC) operating at 77 GHz. Unlike traditional metallic or hybrid counterparts, the proposed AMC is fabricated entirely from a commercial dielectric substrate, offering a low-loss and low-profile solution ideal for integration into compact antenna systems. The proposed AMC demonstrates strong reflection magnitude and achieves a 0° reflection phase at the targeted operating frequency. The −1-dB reflection bandwidth is 3.6 and 8.2 GHz for ± 45° and ± 90° phase windows, respectively. Integration with a dipole antenna confirms that the AMC maintains reflection phase stability under placement and fabrication tolerances. Measured results using a commercial material characterisation kit align closely with simulations, with deviations primarily attributed to substrate anisotropy at 77 GHz. Additionally, two alternative all-dielectric reflectors with high reflection magnitudes and nonzero reflection phases at 77 GHz are presented to support broader millimetre wave scenarios where in-phase reflection is nonessential. This work is among the first to report an all-dielectric AMC operating at millimetre wave frequencies while providing experimental validation, insights into anisotropy effects, fabrication constraints and real-world integration.</p>","PeriodicalId":13374,"journal":{"name":"Iet Microwaves Antennas & Propagation","volume":"19 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/mia2.70039","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144551149","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}

{"title":"A cross dual branch guidance network for salient object detection","authors":"Yiru Wei ,&nbsp;Zhiliang Zhu ,&nbsp;Hai Yu ,&nbsp;Wei Zhang","doi":"10.1016/j.engappai.2025.111480","DOIUrl":"10.1016/j.engappai.2025.111480","url":null,"abstract":"<div><div>The effective integration of multi-level contextual information is crucial for deep learning-based salient object detection. However, most existing approaches either adopt the parallel structure or the progressive structure to predict salient objects, which still face challenges in consistently and accurately detecting salient objects of varying scales. In this paper, we propose a novel cross dual branch guidance network to effectively extract the rich semantic features and gradually enhance the saliency map scale-by-scale. Concretely, the parallel branch is guided by the progressive branch to obtain coarse location information of salient objects. In turn, the progressive branch is able to obtain uniform semantics and rich details to enhance saliency map with the guidance of the parallel branch. To obtain the dynamic receptive field, a dynamic sampling module (DSM) is introduced, which can dynamically adjust the sampling positions such that the spatial details of salient objects in complex scenes can be well recognized. In addition, we design a global context module (GCM) to explore the correlation between different parts of salient object or different salient objects, which is favorable for improving the completeness of saliency map. Experiments on five released benchmark datasets demonstrate the effectiveness and superiority of our proposed approach against other state-of-the-art methods.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"159 ","pages":"Article 111480"},"PeriodicalIF":7.5,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144548783","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}

{"title":"Distribution-aware contrastive learning for domain adaptation in 3D LiDAR segmentation","authors":"Lamiae El Mendili,&nbsp;Sylvie Daniel,&nbsp;Thierry Badard","doi":"10.1016/j.cviu.2025.104438","DOIUrl":"10.1016/j.cviu.2025.104438","url":null,"abstract":"<div><div>Semantic segmentation of 3D LiDAR point clouds is very important for applications like autonomous driving and digital twins of cities. However, current deep learning models suffer from a significant generalization gap. Unsupervised Domain Adaptation methods have recently emerged to tackle this issue. While domain-invariant feature learning using Maximum Mean Discrepancy has shown promise for images due to its simplicity, its application remains unexplored in outdoor mobile mapping point clouds. Moreover, previous methods do not consider the class information, which can lead to suboptimal adaptation performance. We propose a new approach—Contrastive Maximum Mean Discrepancy—to maximize intra-class domain alignment and minimize inter-class domain discrepancy, and integrate it into a 3D semantic segmentation model for LiDAR point clouds. The evaluation of our method with large-scale UDA datasets shows that it surpasses state-of-the-art UDA approaches for 3D LiDAR point clouds. CMMD is a promising UDA approach with strong potential for point cloud semantic segmentation.</div></div>","PeriodicalId":50633,"journal":{"name":"Computer Vision and Image Understanding","volume":"259 ","pages":"Article 104438"},"PeriodicalIF":4.3,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144549098","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":"Semi-supervised generative adversarial networks for imbalanced skin lesion diagnosis with an unbiased generator and informative images","authors":"Mohammad Saber Iraji","doi":"10.1016/j.engappai.2025.111643","DOIUrl":"10.1016/j.engappai.2025.111643","url":null,"abstract":"<div><div>Skin cancer remains a significant global health challenge, necessitating effective early detection methods. Traditional supervised learning approaches for skin lesion classification often require extensive labeled datasets, which are costly and time-consuming to obtain. This study addresses the limitations of semi-supervised learning in skin cancer diagnosis, particularly issues related to classification bias toward majority classes and the impact of incorrect pseudo-labels from low-confidence unlabeled images. We propose an unbiased bad generator within the weighted bad semi-supervised generative adversarial network (WB-SGAN), which integrates self-social consistency regularization and a new weighted inversed cross-entropy loss function for informative images. Additionally, a weighted cross-entropy loss function is used to reduce the bias of the classifier's predictions on labeled and unlabeled images. This framework enhances the generation of informative fake samples, thereby reducing bias in pseudo-labels and improving classification performance. Our experiments demonstrate that WB-SGAN outperforms existing state-of-the-art (SOTA) methods, achieving balanced accuracies of 80.35 % and 74.19 % on the ISIC-2018 and PAD-UFES datasets, respectively, even with just 5 % labeled data. This approach highlights the visual and volumetric aspects of training images and their labeling, offering a solution for skin lesion classification in semi-supervised domains with limited and imbalanced datasets.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"159 ","pages":"Article 111643"},"PeriodicalIF":7.5,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144534452","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}