2890 papers
This paper proposes the Chaotic Improved Cloud Drift Optimization (CI-CDO) algorithm, a novel metaheuristic enhanced by atmospheric turbulence dynamics and chaotic maps, which demonstrates superior performance in solving benchmark optimization problems, selecting optimal feature subsets for classification, and forecasting real-world environmental data.
This paper proposes a metadata-supervised approach to disentangle anatomical structure from acquisition-dependent variations in biomedical imaging, enabling the creation of acquisition-aware representations and a unified harmonisation model that improves generalisation, interpretability, and clinical deployment across diverse imaging protocols and sites.
This study demonstrates that an interpretable ensemble machine learning framework, specifically utilizing Extra Trees Regression, can accurately predict the diameter of TiO₂–PVP electrospun nanofibers based on process parameters, thereby offering a highly efficient and precise alternative to traditional time-consuming experimental optimization methods.
This paper proposes MSF-TrashDet, a lightweight multi-scale feature fusion network that utilizes a novel Multi-Scale Feature Enhancement block and structured channel pruning to achieve high-accuracy, efficient waste object detection on the TrashNet dataset, outperforming state-of-the-art models like YOLOv5n, YOLOv8n, and YOLOv11n with 91.91% mAP50 while maintaining low computational complexity.
This paper proposes a scalable hybrid encryption framework that combines the Unimodular Hill Cipher with a chaotic Logistic Map-based XOR stream cipher to provide secure, efficient, and format-independent protection for arbitrary binary files, achieving near-ideal randomness, strong resistance to statistical attacks, and lossless recovery with high throughput.
This paper presents the Hybrid Split-Based Chaotic Cryptosystem (HSBCC), a unified framework that integrates a dynamically generated unimodular Hill Cipher, Logistic Map-based key derivation, and Elliptic Curve Cryptography to achieve secure, lossless, and size-preserving encryption for arbitrary binary files.
This study demonstrates that while block-sparse pruning significantly compresses a deep learning intrusion detection model for edge deployment without improving inference latency on standard hardware, it introduces class-specific instability and fails to resolve performance issues for minority attack types.
This paper proposes an AI-driven framework for federated cloud environments that integrates graph neural networks, NSGA-II optimization, and contextual bandits to simultaneously minimize SLO violations, energy consumption, operational costs, and inter-cluster traffic while maximizing fairness and stability, achieving superior performance over existing methods in experimental evaluations.
This study evaluates six conversational AI platforms and finds that while they consistently simulate empathy, they exhibit significant inconsistencies in crisis detection and escalation, revealing a critical gap between their documented safety commitments and actual interactional performance.
This paper proposes a unified two-phase hybrid framework that integrates a regime-aware motion classification pipeline with reinforcement learning-based adaptive tuning of Kalman filter noise parameters to enhance autonomous vehicle navigation accuracy and robustness against GPS spoofing.