Computers

This paper reviews sensor technologies within IoT frameworks, detailing their architecture, classifications, and roles in wireless systems. It examines recent trends and research challenges to guide future innovation for the industry: mdpi.com/2073-431X/14/8…

This paper presents an #AI-driven system using #COMSOL models and IR thermography to monitor #PCB thermal stress. Combining U-Net and #MLP architectures, it achieves high accuracy in detecting anomalies for biomedical device maintenance: mdpi.com/2073-431X/14/9…

This paper breaks down complex concepts into digestible insights, prioritizing clarity and precision. By synthesizing key data and trends, it offers a grounded perspective on the subject, ensuring the reader gains a comprehensive understanding quickly: mdpi.com/2073-431X/15/3…

This paper addresses the problem of multi-class classification of bacterial microscopic images using a rigorous experimental protocol designed to prevent information leakage and improve performance: mdpi.com/2073-431X/15/3…

This paper evaluates XRSkills, a VR serious game designed to help teachers manage classroom stress. Results show good usability and engagement, suggesting that immersive simulations are a viable, safe way for educators to practice coping strategies: mdpi.com/2073-431X/15/3…

This paper introduces TrustGTN, a graph neural network using soft selection to learn trust patterns in social networks. By capturing evolving structural and semantic data without manual tuning, it outperforms traditional trust evaluation methods: mdpi.com/2073-431X/15/3…

This paper introduces ExamQ-Gen, an instructor-in-the-loop benchmark for LLM grading. It finds that while score agreement is strong, threshold-based grading suffers from systematic conservatism, leading to high rates of false failures for students: mdpi.com/2073-431X/15/3…

This paper reviews fog computing IDSs via PRISMA. Analyzing 4,905 papers (2021–2025), it finds hybrid/deep learning models achieve 95–99% accuracy but face latency and privacy gaps. Future focus: lightweight, federated, and real-world evaluation: mdpi.com/2073-431X/15/3…

REMINDER! Only 1 Month Left to Submit Your Abstracts for the 3rd International Conference on AI Sensors and Transducers. Submit Your Abstracts by 31 March 2026: sciforum.net/user/submissio…

This Special Issue focuses on advances that bridge AI/ML, security automation, and adaptive architectures to build proactive, scalable, and trustworthy cyber defense systems. Learn more by visiting the webpage: mdpi.com/journal/comput…

This Special Issue aims to bring together cutting-edge research that spans the full spectrum of complex data intelligence, from efficient retrieval and representation learning to semantic analysis, reasoning, and decision-making applications: mdpi.com/journal/comput…

This paper reviews smart precision irrigation as a solution to water scarcity. It examines sensors, aerial systems, AI, and agroclimatic models to optimize water use. By highlighting benefits and limits, it guides sustainable, efficient farming: mdpi.com/2073-431X/15/2…

This paper details a hybrid CAT model merging neural networks with IRT to fix the "cold-start" problem. Using student data for early estimates, it cut initial error by 34.2% and shortened tests by 28.7% while maintaining high precision: mdpi.com/2073-431X/15/2…

In this recently published paper authors design, implement and evaluate a trustworthy, blockchain-based food supply chain traceability system equipped with an incentive mechanism: mdpi.com/2073-431X/15/2…

This paper evaluates open-source LLMs like Gemma 2 and Phi-3 for coding tasks using SVM and RankNet. It uses metrics like code complexity and ROUGE scores to move beyond subjective selection, achieving up to 90% precision in model ranking: mdpi.com/2073-431X/15/2…

This paper introduces Trust-FedGNN, a framework for IoT that boosts accuracy by 5.8% and cuts energy use by 22%. It uses reputation modeling and graph pruning to secure distributed learning while protecting resource-constrained devices: mdpi.com/2073-431X/15/2…

This paper proposes ConvLSTM and ConvBiLSTM models using linear interpolation data augmentation to forecast dengue cases in Peru. These models outperform standard RNNs like LSTM and GRU, offering higher accuracy with lower computational costs: mdpi.com/2073-431X/15/2…

This paper details a new lung cytology model using Vision Transformers and LLMs to generate imaging findings. Validated on 1059 images, it shows superior accuracy and classification, proving task-specific models enhance diagnostic support: mdpi.com/2073-431X/15/2…

The editorial office of Computers would like to extend its sincere gratitude to all reviewers who contributed to the improvement of the journal quality by providing their expert opinion and evaluation of the submitted research: mdpi.com/journal/comput…

This paper presents a blockchain-based food traceability system using Hyperledger Fabric. It features a multi-channel architecture for privacy, plus trust and incentive mechanisms to ensure data credibility and stable, scalable performance: mdpi.com/2073-431X/15/2…