Zhuodong Liu¹, Xiangyu Li^2,*, Zhihao Zhang¹

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.085409 - 23 July 2026

Abstract Federated unlearning (FU) enables the removal of specific data contributions from federated learning (FL) models to comply with regulations such as the General Data Protection Regulation (GDPR). However, most existing FU methods are designed for the FedAvg paradigm, where all clients share a single global model. In practice, personalized federated learning (pFL) methods such as FedPer, FedRep, Ditto, and FedBN have become widely adopted due to their superior handling of non-IID data. These methods decompose the model into shared global layers and client-specific personalized layers, fundamentally altering the semantics of unlearning, yet this setting has… More >

Mujahid Ali^*

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.084555 - 23 July 2026

Abstract Understanding the determinants of travel mode choice (TMC) in urban contexts is essential for effective transport planning and policy development. Past studies predominantly employed traditional discrete choice models because of their simplicity, diversity, and high interpretability; however, they rely on restrictive assumptions. Although machine learning (ML) techniques have shown promising predictive capabilities, comparative assessments of traditional and ML approaches, particularly considering hyperparameter optimisation, remain limited. This study addresses this gap by comparing a traditional model with four ML algorithms: decision tree (DT), random forest (RF), support vector machine (SVM), and k-nearest neighbour (KNN). In addition,… More >

Jing-Wen Gao^1,2, Yunan He^1,*, Jian Liu^1,*

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.084503 - 23 July 2026

Abstract The unique topological properties of the electronic band structures in topological materials have increasingly attracted attention in both fundamental research and next-generation technological applications. With the rise of machine learning, the connection between topological materials and machine learning has deepened significantly. This review systematically summarizes the interaction between these two fields, tracing the history of their mutual promotion and synergistic development. We further examine the transformative impact of machine learning across multiple domains of topological materials, with a particular focus on recent progress in inverse design and generation of topological materials, topological superconductivity, and the More >

Yixiang Wan, Wenqiu Zhu^*

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.084474 - 23 July 2026

Abstract Addressing the two critical challenges of missed detection of distant small targets and difficulty in identifying occluded targets under complex road conditions, this paper proposes YOLO-PBE, an improved high-precision vehicle detection model based on YOLOv11n. First, to tackle the fine-grained feature loss caused by conventional strided convolutions during downsampling, we add a high-resolution P2 detection layer and introduce SPD-Conv, a lossless spatial-to-depth feature transformation technique, for feature extraction. By preserving complete pixel-level information, the model's perception accuracy for distant small vehicles is enhanced. For feature fusion, we design an improved BiFPN incorporating a Ghost module.… More >

Mubariz Khan¹, Hafeez Ur Rehman Siddiqui², Adil Ali Saleem², Muhammad Amjad Raza^2,3, Lázaro Javier Hernández Rodríguez^4,5,6,7, Pablo Herrero García^4,8,9, Isabel de la Torre Díez^10,*

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.084269 - 23 July 2026

Abstract Accurate forecasting of cryptocurrency prices remains an open challenge because classical statistical models cannot capture the non-linear, sentiment-driven dynamics of these markets. This study compares three hybrid deep learning architectures—VAR-LSTM, XGBoost-LSTM, and CNN-LSTM—to determine which best forecasts Bitcoin (BTC), Ethereum (ETH), and Dogecoin (DOGE) closing prices, and to quantify the marginal predictive value of Twitter sentiment integration. Six years of hourly OHLCV data (2017–2023) are augmented with VADER-scored Twitter sentiment polarity. Each model is formulated mathematically, implemented with documented hyperparameters (epochs, dropout, units;), and trained for one-step-ahead next-hour price prediction. Performance is measured by RMSE,… More >

Xing Fang¹, Yuanfang Chen^1,2,*, Qiang Lin³, Kun Yang^2,4, Gyu Myoung Lee⁵

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.084208 - 23 July 2026

Abstract The deployment of supervised anomaly detection is typically limited by the high cost of annotation, privacy constraints, and the scarcity of anomalous samples. These constraints have motivated the use of vision-language pre-trained models for zero-shot anomaly detection. However, existing CLIP-based methods still face three limitations: a shared set of prompts is applied across feature layers, anomaly maps are fused by fixed strategies, and image-level anomaly scores are determined solely by global image-text similarity. These limitations reduce the accuracy of pixel-level localization and weaken the reliability of image-level anomaly prediction. To overcome these limitations, LaRP-CLIP is More >

Yun Liu¹, Jinghua Zhao¹, Liang Ma¹, Zijie Huang^2,3,*, Lizhi Cai^2,3, Jianxin Ge^2,3

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.084179 - 23 July 2026

Abstract Automated library migration reduces refactoring costs but challenges traditional evolutionary algorithms, which often suffer from premature convergence and poor recall in sparse, complex API mapping spaces. To address this, we propose QIMIG, a multi-objective optimization framework integrating quantum-inspired encoding with quality-aware and greedy heuristic filtering. QIMIG utilizes a probabilistic Q-bit representation to maintain population diversity and avoid local optima. Simultaneously, its heuristic components leverage historical usage context to filter semantic noise and guide the search toward valid mappings. Evaluated on 9 real-world migration rules derived from 57,447 open-source projects, QIMIG statistically significantly outperforms state-of-the-art baselines More >

Lijun Yang¹, Qiuwei Yang^2,*

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.083976 - 23 July 2026

Abstract This study proposes a Multi-Stage Sparrow Search Algorithm (MS-SSA) for precise structural damage identification. Initially, the structural static displacement sensitivity formulation is derived via the Sherman-Morrison-Woodbury formula, and an objective function is constructed by integrating the sensitivity equations with the L2-norm penalty. Subsequently, MS-SSA is implemented to sequentially achieve preliminary damage localization and accurate quantification. In the localization phase, a constrained narrow-bound search space is predefined to identify potential damage regions. Leveraging this feedback, the sensitivity equations are condensed, and the search boundaries are adaptively refined for the quantification phase, where SSA is reapplied to… More >

Osama Al-Haj Hassan^1,*, Ammar Odeh¹, Abdullah Aref ², Ghassan Samara³

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.083861 - 23 July 2026

Abstract Mashups are among the key web technologies that provide end-users with customizable and personalized tools. Most mashup platforms are based on centralized architectures or do not employ fully decentralized architectures; therefore, in this paper, we propose a decentralized architecture for mashups that combines the strengths of structured and unstructured peer-to-peer networks. For the structured part, we rely on the Chord lookup protocol, and for the unstructured part, we build groups of nodes via two flavors of network flooding, namely, sequence number flooding and reverse path flooding. Brokers in the unstructured part would be responsible for More >

Amir Ijaz^*, Hashem Haghbayan, Abdul Malik, Ethiopia Nigussie, Juha Plosila

CMC-Computers, Materials & Continua, Vol.88, No.3, 2026, DOI:10.32604/cmc.2026.083797 - 23 July 2026

Abstract The exponential growth of the Internet of Things (IoT) has led to an urgent need for highly energy-efficient communication strategies, especially for battery-powered or self-sustaining devices. In this work, we present a comprehensive framework for minimizing communication energy in IoT nodes operating in swarm robotic systems. We examine and integrate multiple low-power wireless technologies (BLE, LoRaWAN, MQTT, CoAP) with advanced Medium Access Control (MAC) protocols. We additionally propose adaptive scenarios leveraging both ambient energy harvesting and passive backscatter transmission. Our solution employs adaptive scheduling and dynamic transmission power management. Specifically, a Deep Q-Learning (DQL) agent More >