SIGNRL: A Population-Based Reinforcement Learning Method for Continuous Control.
Daniel F. Zambrano-GutierrezAlberto C. Molina-PorrasEmmanuel Ovalle-MagallanesIván AmayaJosé Carlos Ortiz-BaylissJuan Gabriel Aviña-CervantesJorge M. Cruz-DuartePublished in: SSCI (2023)
Keyphrases
- high accuracy
- cost function
- reinforcement learning
- experimental evaluation
- significant improvement
- high precision
- genetic algorithm
- detection method
- synthetic data
- segmentation method
- computational complexity
- preprocessing
- pairwise
- support vector
- similarity measure
- image segmentation
- classification accuracy
- particle swarm optimization
- machine learning