Transactions on Machine Intelligence2821-16938220250601Resource Management in Vehicular Fog Networks Based on Contract Theory697924484110.22034/tmi.2025.244841ENR.HaghighizadehDepartment of Electrical Engineering, School of Electrical and Computer Engineering, Isfahan University of Technology, Isfahan, IranZ.ZaliDepartment of Electrical Engineering, School of Electrical and Computer Engineering, Isfahan University of Technology, Isfahan, IranJournal Article20241122Fog computing is a distributed infrastructure that extends computing, communication, and storage capabilities toward the network edge. Compared to cloud computing, fog computing can support delay-sensitive service requests while reducing energy consumption and traffic congestion. Fog computing contributes to efficient resource utilization and improved performance in terms of latency, bandwidth, and energy consumption. However, one of the major challenges in fog computing is the limited computational capacity of fog nodes under increasing daily demand, especially during peak hours, which can lead to severe performance degradation. Therefore, an optimal mechanism is required to ensure satisfactory quality of service (QoS). Integrating fog computing with vehicular ad hoc networks, leading to vehicular fog computing (VFC), has emerged as a promising solution to reduce overload at base stations and minimize processing delays during peak periods. In this approach, the surplus computational resources of nearby vehicles are utilized as an on-demand, low-cost option. Consequently, the computing resources provided by a large group of vehicles can be aggregated to alleviate network congestion during peak hours without additional servers, enabling real-time computational scalability. Nevertheless, the large-scale deployment of vehicular fog networks still faces several critical challenges, such as the lack of efficient incentive mechanisms and task assignment strategies. In this study, we first propose a solution to minimize network delay from the perspective of integrated contract-based optimization. Next, the problem of computational task allocation is formulated as a two-sided matching problem between vehicles and users, and a stable, QoS-aware matching algorithm is introduced to solve it. Finally, task offloading decisions are performed to minimize total network latency. The proposed scheme can effectively guarantee network load balancing and improve the utilization of idle vehicular resources.Transactions on Machine Intelligence2821-16938220250601Dynamic Modeling of Power Systems in the Presence of Solid Oxide Fuel Cells and Wind Turbines809024484210.22034/tmi.2025.244842ENJournal Article20241226In this paper, the small-signal stability of a power system is investigated when a Solid Oxide Fuel Cell (SOFC) power plant and a Doubly Fed Induction Generator (DFIG)-based wind turbine operate in parallel with a synchronous generator in the power grid. First, a comprehensive mathematical model of the SOFC generator and the DFIG-based wind turbine integrated into a single-machine infinite-bus power system is developed. The derived mathematical models are then linearized. Based on the linearized model of the overall system, the Heffron–Phillips model corresponding to the proposed configuration is obtained, and subsequently, small-signal stability analysis of the power system is carried out. The results demonstrate that the presence of wind power generation and SOFC units may influence system stability both positively and negatively, depending on different system parameters and operating conditions.Transactions on Machine Intelligence2821-16938220250601Application of the Harmony Search Metaheuristic Algorithm in Distribution Network Optimization with Focus on Reconfiguration and Simultaneous Installation of Distributed Generation Units819024484310.22034/tmi.2025.244843ENA.Abdi FarzaniMaster’s Student, Department of Electrical Engineering, Hakim Sabzevari University, Sabzevar, IranR.RoshanfekrAssistant Professor, Department of Electrical Engineering, Hakim Sabzevari University, Sabzevar, IranJournal Article20250111<br>Distribution network reconfiguration and the optimal utilization of distributed generation (DG) resources are effective methods for reducing losses and improving power quality in electrical distribution systems. In recent years, the use of DG has attracted significant attention due to its notable advantages. Integrating DG can reduce network losses and enhance voltage stability. Furthermore, the use of DG resources contributes to increasing the reliability of the system. Therefore, optimal placement of these resources, combined with appropriate distribution network reconfiguration, can have a substantial impact on overall system performance. In this study, a novel approach for reconstructing the distribution network structure in the presence of DG units is proposed. The primary objective of this approach is to minimize power losses and bus voltage deviations while improving the voltage profile in the distribution system. To achieve these objectives, the Harmony Search Algorithm (HSA), a metaheuristic optimization method, is employed for network reconfiguration and simultaneous identification of optimal DG placement. Additionally, sensitivity analysis is used to determine the most suitable locations for DG installation. The proposed method is tested on a 33-bus radial distribution system to demonstrate its efficiency and effectiveness. The results obtained from this study are satisfactory.Transactions on Machine Intelligence2821-16938220250601Brain MRI Segmentation Using an Improved Bat Optimization Algorithm9110024485010.22034/tmi.2025.244850ENW.RenhuaDepartment of Electronic Engineering, Tsinghua University, Beijing, ChinaJournal Article20250203Magnetic Resonance Imaging (MRI) is a non-invasive diagnostic modality extensively utilized in clinical medicine. In the computational analysis of brain MRI scans, image segmentation yields critical spatial knowledge regarding internal anatomical structures. Although this segmentation process is conventionally performed manually by radiologists, its precise execution is hindered by the complex morphology of soft tissues and the presence of system-induced artifacts or noise. This study introduces an automated image segmentation framework leveraging an Improved Bat Optimization Algorithm. Under this approach, the bat heuristic is hybridized with the $K$-means algorithm to globally optimize the selection of initial cluster centroids. The operational efficiency of the proposed method is evaluated and benchmarked against alternative state-of-the-art techniques. Quantitative simulations executed within the MATLAB environment demonstrate that the proposed framework achieves an outstanding segmentation accuracy of 99.5%, consistently outperforming baseline methods.Transactions on Machine Intelligence2821-16938220250601Leveraging Bagging Ensemble Architectures for the Automated Diagnosis of Dyslexia via Visual Task Paradigm Analysis10111024485110.22034/tmi.2025.244851ENM.ZareeDepartment of Biomedical Engineering, K. N. Toosi University of Technology, Tehran, IranB.Arefi-RadDepartment of Electrical Engineering, K. N. Toosi University of Technology, Tehran, IranM.MohebbiDepartment of Biomedical Engineering, K. N. Toosi University of Technology, Tehran, IranR.RostamiDepartment of Psychology, University of Tehran, Tehran, IranJournal Article20250213Dyslexia is a neurobiological learning disability that fundamentally impairs a child’s literacy acquisition, specifically manifesting as persistent deficits in reading and writing. Absent a timely diagnosis, the disorder precipitates profound psychological distress and academic marginalization for both the pediatric patients and their families. Furthermore, delayed intervention often results in cumulative achievement gaps that become increasingly difficult to bridge by secondary education. Consequently, early screening and clinical intervention are paramount to preserving student self-esteem and optimizing long-term academic trajectories. This study proposes an automated diagnostic framework utilizing a Bagging (Bootstrap Aggregating) ensemble learning approach to classify dyslexia in children. The methodology involves the rigorous preprocessing of electroencephalogram (EEG) signals recorded across a 19-channel montage. Feature extraction focused on the morphometry of Event-Related Potentials (ERPs), specifically quantifying the amplitude and latency of key components. To address the "curse of dimensionality" inherent in the high-dimensional feature set, Principal Component Analysis (PCA) was implemented for optimal feature reduction. To ensure the generalizability of the model and mitigate the risk of overfitting, a K-fold cross-validation strategy was employed during the training phase. Finally, the Bagging classifier was deployed to distinguish between dyslexic and neurotypical subjects. The proposed ensemble framework demonstrated robust performance, yielding an average classification accuracy of 90.6%. Notably, the model achieved a sensitivity rate of 100%, ensuring no dyslexic cases were omitted, and a specificity of 81.2%, reflecting its capability to accurately identify neurotypical controls.Transactions on Machine Intelligence2821-16938220250601Medical Image Denoising Using the Cuckoo Optimization Algorithm11112024485310.22034/tmi.2025.244853ENA.AliabadianAssistant Professor, Department of Electrical and Biomedical Engineering, Shomal University0000-0002-8646-7968Journal Article20250204Medical images acquired through modalities such as MRI, CT, and ultrasound are inherently corrupted by various types of noise during acquisition and transmission, which severely complicates accurate clinical diagnosis and quantitative interpretation. The bilateral filter has emerged as a widely accepted and powerful framework for edge-preserving noise reduction due to its unique capability to simultaneously utilize both spatial domain and intensity range information. However, dynamically determining its optimal parameters and window sizes for different noise levels remains a persistent challenge in image processing. To address this limitation, this study introduces a robust, modified bilateral filter whose weights and parameters are automatically and intelligently optimized using the Cuckoo Optimization Algorithm (COA). Utilizing an adaptive global search mechanism inspired by the parasitic brooding behavior of cuckoos, the proposed algorithm successfully identifies the optimal combination of filter coefficients by maximizing an objective fitness function based on image quality metrics. Experimental evaluations on benchmark medical datasets demonstrate that the proposed hybrid method significantly outperforms conventional filtering techniques. The quantitative and qualitative results indicate that this approach not only effectively suppresses noise under high-density scenarios but also remarkably preserves critical structural details, fine textures, and sharp boundaries, thereby enhancing the diagnostic value of the clinical images.