Enhancing Cloud Security: A Study on Ensemble Learning-Based Intrusion Detection Systems
The study focuses on enhancing cloud security using ensemble learning algorithms. Ensemble RUSBoost achieved the best accuracy at 99.821% in…
Advancements in Robotic Seafloor Mapping: A Navigation-Aided Hierarchical Approach
Underwater mapping faces challenges from light attenuation and scattering. Hierarchical reconstruction integrates SLAM and global SfM for completeness. Extensive testing…
A Wireless Drive and Control Method for Robots: Multifrequency Microwaves
New method for wireless robot control using multifrequency microwaves introduced. Millimeter-scale robot designed using shape memory alloy for confined environments.…
Vision-Based Approach Enhances Deformable Robot Key Point Estimation
New vision-based method improves deformable robot control accuracy by up to 4.5%. Utilizes two cameras and a convolutional neural network…
Stretchable Thermoelectric Generators for Self-Powered Wearable Health Monitoring
Stretchable TEGs offer sustainable power for wearable health devices. Combines 3D printed elastomers with liquid metal composites. Powers PPG monitoring…
Two-Layer Shipboard Energy Management Framework Utilizing Reinforcement Learning
Shipboard energy management requires integrating navigation planning and real-time energy control. A two-layer framework using particle swarm optimization and deep…
Anomaly Detection in Surveillance Videos Enhanced by Mutual Learning and Cropped Snippets
Mutual learning improves detection accuracy in surveillance videos. Cropped snippets mitigate noise, enhancing feature training. Achieved 85.78% frame-level AUC on…
A Novel Ensemble Deep Reinforcement Learning Model for Short‐Term Load Forecasting Based on Q‐Learning Dynamic Model Selection
The proposed model uses Q-learning for dynamic weight adjustment. RNN, LSTM, and GRU are the primary predictors in the ensemble.…
AI-Based Method for Improving Frequency Response Testing and Measurement Error Prediction in DCTV
AI-based method enhances DCTV frequency response tests and error prediction. Combines AC and DC superposition, voltage change, and phase correction.…
Computer Vision Pioneers in Detecting Eye Diseases Using Pre-trained ImageNet Models
The study uses pre-trained ImageNet models for automated eye disease detection. InceptionResNetV2 achieved the highest accuracy of 93% among tested…
