On Selecting Optimal Hyperparameters for Reinforcement Learning Based Robotics Applications: A Practical Approach.
Ignacio Fidalgo AstorquiaGuillermo VillateAlberto TellaecheJuan Ignacio VázquezPublished in: ICINCO (1) (2023)
Keyphrases
- hyperparameters
- reinforcement learning
- closed form
- model selection
- cross validation
- dynamic programming
- bayesian framework
- support vector
- maximum likelihood
- bayesian inference
- em algorithm
- gaussian process
- computer vision
- regularization parameter
- gaussian processes
- random sampling
- machine learning
- incremental learning
- maximum a posteriori
- prior information
- sample size
- noise level
- state space
- pairwise
- learning environment
- feature selection
- learning algorithm
- data sets