Adaptive Optimization of Hyper-Parameters for Robotic Manipulation through Evolutionary Reinforcement Learning.
Giulio OnoriAsad Ali ShahidFrancesco BraghinLoris RovedaPublished in: J. Intell. Robotic Syst. (2024)
Keyphrases
- hyperparameters
- reinforcement learning
- model selection
- parameter optimization
- cross validation
- random sampling
- support vector
- bayesian inference
- manipulation tasks
- bayesian framework
- prior information
- grid search
- genetic algorithm
- dynamic programming
- maximum likelihood
- noise level
- sample size
- closed form
- machine learning
- supervised learning
- incremental learning
- prior knowledge
- learning algorithm
- variational bayes
- em algorithm
- edge detection
- genetic algorithm ga
- incomplete data
- computer vision