Transactions on Machine Intelligence2821-16938320250901Modeling the Impact of Soil Liquefaction on Structural Stability Using an Artificial Neural Network Optimized by the Cuckoo Optimization Algorithm12113024485510.22034/tmi.2025.244855ENM.Foroutan NaddafiDepartment of Civil Engineering and Architecture, National University of Skills, Tehran, Iran0000-0002-3947-6423Journal Article20250403Soil liquefaction is one of the most critical geotechnical phenomena that can severely impact the stability and performance of engineering structures during seismic events. Accurate prediction of liquefaction potential and its subsequent effects on structural stability is a complex, non-linear problem influenced by a combination of intertwined geotechnical and seismic parameters. In this research, an Artificial Neural Network (ANN) model was developed to simulate and predict the impact of soil liquefaction on structural stability using a comprehensive dataset of geotechnical and seismic features. To enhance the predictive performance and generalization capability of the neural network, the hyperparameters and network architecture were optimized using the Cuckoo Optimization Algorithm (COA) an algorithm that enables the simultaneous optimization of multiple conflicting objectives, such as prediction accuracy and model complexity. The optimized neural network demonstrated highly superior performance in classifying liquefaction and non-liquefaction states, delivering high accuracy and remarkable stability on the validation dataset. Furthermore, the hybrid ANN–COA framework provides a reliable, efficient, and computational approach for assessing the impact of liquefaction on structural stability, offering valuable insights for seismic design, risk assessment, and the formulation of hazard mitigation strategies in geotechnical engineering.Transactions on Machine Intelligence2821-16938320250901Detection of Atrial Fibrillation via Analysis of Mean, Entropy, and Heart Rate Parameters Using Electrocardiogram Signals12113024486010.22034/tmi.2025.244860ENM.AghaeiM.Sc. Student in Biomedical Engineering (Bioelectric Track), Department of Biomedical Engineering, Semnan University, Semnan. IranJournal Article20250304Atrial Fibrillation (AF), commonly referred to as AFib in clinical medicine, is the most prevalent type of cardiac arrhythmia. Atrial Fibrillation manifests when the electrical excitation wave propagates through the atria without a defined spatial direction or coordinated pathway. This paper analyzes the electrocardiogram (ECG) signals of adults, benchmarking healthy individuals against those diagnosed with Atrial Fibrillation. The dataset utilized in this study was compiled by the American and European Heart Associations. Regarding the demographic distribution, less than 5% of the subjects fall within the 40–50 age bracket, while 5–15% are in the octogenarian category (around 80 years old). Based on the quantitative investigations, the heart rate in patients suffering from Atrial Fibrillation exhibits highly irregular dynamics and elevated magnitudes, consistently clustering in the range of 80 to 100 beats per minute (bpm), and frequently exceeding 100 bpm. Furthermore, statistical feature extraction reveals that individuals with Atrial Fibrillation present a significantly lower mathematical mean, alongside markedly higher entropy and energy values in their processed ECG signal segments compared to healthy cohorts.Transactions on Machine Intelligence2821-16938320250901Speed Control Optimization of Permanent Magnet Synchronous Motors (PMSM) Using the Cuckoo Optimization Algorithm (COA)13114024486110.22034/tmi.2025.244861ENM.MoradiDepartment of Electrical and Electronics Engineering, Shuhada-e-Hoveyzeh University of Technology, Shahid Chamran University of Ahvaz, Ahvaz, IranM.MoghadasianAssistant Professor, Department of Electrical and Electronics Engineering, Shahid Chamran University of Ahvaz, Ahvaz, IranJournal Article20250423This paper presents the speed control optimization of a Permanent Magnet Synchronous Motor (PMSM) utilizing the Cuckoo Optimization Algorithm (COA). Due to their high power density, superior efficiency, and robust structural design, PMSMs are extensively utilized in a wide range of industrial and domestic applications. However, achieving precise and optimal speed control in these motors remains a significant challenge due to their non-linear dynamics. In this research, the parameters of a Proportional-Integral-Derivative (PID) controller are optimized to minimize speed tracking errors and enhance the motor's dynamic response. The Cuckoo Optimization Algorithm (COA) was selected for this optimization task owing to its algorithmic simplicity, fast convergence, and high efficiency in solving complex engineering problems. The problem formulation incorporates an objective function designed based on the Integral Squared Error (ISE) criterion, where the decision variables are defined as the tuning gains of the PID controller. Simulation studies conducted in MATLAB/Simulink R2024a demonstrate that employing the COA for PID parameter tuning yields a substantial improvement in the performance of the PMSM drive system. Specifically, a significant reduction in speed error, suppressed overshoot, and an accelerated system settling time are achieved using the proposed optimization approach. Finally, the proposed method is benchmarked against other metaheuristic optimization techniques, confirming its superior performance and efficacy.Transactions on Machine Intelligence2821-16938320250901ECG Signal Compression in Wireless Body Area Networks Using the Bat Algorithm14115024486210.22034/tmi.2025.244862ENP.KeshaniDepartment of Computer Science and Automation, Indian Institute of Science, Bengaluru, IndiaJournal Article20250604In this research, a novel method for Electrocardiogram (ECG) signal compression is proposed based on the Discrete Wavelet Transform (DWT) and the Bat Algorithm. In this approach, the raw signal is transformed into the wavelet domain, where the Bat Algorithm is employed to select the optimal wavelet coefficients by maximizing the Peak Signal-to-Noise Ratio (PSNR) metric. The selected coefficients capture the primary information of the signal, while the remaining coefficients are discarded to achieve a high compression ratio while maintaining reconstruction quality. Performance evaluations conducted on the MIT-BIH database demonstrate that the proposed method exhibits a distinct superiority in terms of PSNR compared to conventional techniques. The results indicate that this method can serve as an effective, low-power solution for real-time ECG signal compression in Wireless Body Area Networks (WBANs).Transactions on Machine Intelligence2821-16938320250901Multi-Objective Genetic Algorithm-Based Clustering for Energy Distribution and Lifetime Maximization in Wireless Sensor Networks15116024486310.22034/tmi.2025.244863ENA.AliabadianAssistant Professor, Department of Electrical and Biomedical Engineering, Shomal University0000-0002-8646-7968Journal Article20250610Wireless Sensor Networks (WSNs) consist of numerous resource-constrained sensor nodes that collaboratively monitor physical or environmental conditions and transmit the collected data to a base station. Due to the limited battery capacity of sensor nodes, energy efficiency remains one of the most critical challenges affecting network lifetime and overall performance. Clustering has been widely adopted as an effective approach for reducing communication overhead, balancing energy consumption, and improving network scalability. However, traditional clustering protocols such as LEACH often suffer from inefficient cluster-head selection and uneven energy distribution among sensor nodes. To address these limitations, this paper proposes a clustering methodology based on a Multi-Objective Genetic Algorithm (MOGA) that simultaneously optimizes cluster-head selection, intra-cluster communication distance, and energy utilization. The proposed approach aims to achieve balanced energy consumption while extending network lifetime and maintaining communication reliability. Extensive simulations were conducted and the obtained results were compared with conventional clustering techniques. Performance evaluation demonstrates that the proposed method significantly reduces energy consumption, improves cluster stability, and increases overall network efficiency. The results indicate that the proposed optimization framework provides a robust and effective solution for energy-aware clustering in WSN environments.Transactions on Machine Intelligence2821-16938320250901Deep Learning and Bat Algorithm-Based Human Activity Recognition16117024486410.22034/tmi.2025.244864ENV.RojasDepartment of Electrical Engineering, Pontificia Universidad Católica de Chile, Santiago, ChileJournal Article20250505Human Activity Recognition (HAR) is a rapidly evolving research area that focuses on identifying and classifying human activities using sensor- and vision-based data. HAR has gained significant attention due to its wide range of applications in intelligent surveillance systems, biometric identification, smart environments, healthcare monitoring, and human–computer interaction. Accurate and real-time activity recognition is particularly important in surveillance applications, where the timely detection of suspicious behaviors can contribute to crime prevention and public safety. Recent advances in deep learning have demonstrated remarkable performance in HAR tasks, especially through Convolutional Neural Networks (CNNs), which are capable of automatically extracting discriminative features from visual data. However, the performance of CNN-based models is highly dependent on the selection of optimal network parameters. To address this challenge, this paper proposes a hybrid HAR framework that integrates CNNs with the Bat Optimization Algorithm (BOA) to enhance feature learning and classification performance. The proposed method is evaluated using the Weizmann human activity dataset and compared with several existing HAR approaches. Experimental results demonstrate that the integration of BOA with CNN improves recognition accuracy and classification effectiveness, highlighting the potential of the proposed framework for intelligent activity recognition applications.