833 papers
This paper proposes RL-Trust RPL, a Reinforcement Learning-based adaptive trust framework that mitigates sinkhole attacks in Multicast RPL networks by monitoring node behaviors like packet delivery and energy consumption to detect and isolate malicious nodes, thereby enhancing routing security and network performance.
This paper proposes a hybrid Genetic Algorithm and Deep Learning framework that integrates a lightweight ResNet model with a custom GA-driven optimization scheme to enable rapid, energy-efficient cluster head selection in Rechargeable Wireless Sensor Networks, thereby significantly delaying the First Node Death and achieving near real-time performance without the iterative overhead of traditional heuristic methods.
This study proposes a scalable, big data-driven framework utilizing a Spatio-Temporal Parallel Transformer (STPT) and clustering techniques to accurately predict traffic accidents and fatalities in Bangkok, thereby enabling evidence-based policy interventions to enhance urban road safety.
This paper proposes PHUE, a progressive hardness-aware undersampling ensemble method that dynamically selects critical majority-class samples near the decision boundary and employs multi-metric weighted integration to significantly outperform existing algorithms in imbalanced data classification.
This paper proposes the Moment-guided Progressive Geometric Reasoning (MPGR) framework, which addresses robust 6D pose estimation under severe occlusion by reformulating the task as a distribution-aware inference problem that utilizes high-order statistical modeling to repair fragmented geometric structures and enhance feature consistency.
This paper proposes a graph neural network-based framework that models team sports as dynamic heterogeneous graphs to predict offensive efficiency and defensive formations while optimizing tactical strategies through graph reinforcement learning and counterfactual reasoning, demonstrating superior performance over existing baselines on UEFA Champions League and NBA datasets.
This paper evaluates YOLOv11 and EfficientNetV2M models on the imbalanced HAM10000 dataset for skin tumor classification, demonstrating that YOLOv11 achieves superior recall (86.5%) for high-sensitivity screening while EfficientNetV2M combined with an SVM classifier offers higher specificity (96.8%) for precise triage.
This paper proposes a comprehensive hybrid quantum-classical framework comprising five specialized models (QAFEN-CNN, VQTM-LSTM, QGCL-Net, QVAE-DD, and QMRL-XAI) that collectively enhance financial fraud detection accuracy, adaptability to dynamic ecosystems, and regulatory explainability through advanced quantum feature entanglement, graph construction, and meta-reinforcement learning techniques.
This survey synthesizes emerging research on autonomous vulnerability remediation across diverse computing environments, organizing the literature into six key streams and a formal closed-loop task model to highlight critical themes, operational safety mechanisms, and significant research gaps in deploying agentic AI and reinforcement learning for secure system repair.
This paper establishes that the per-task energy and water demands of AI inference are orders of magnitude lower than those of the human labor it augments or replaces, even when accounting for the full lifecycle resource footprint of human work.