Reinforcement learning to adjust parametrized motor primitives to new situations.
Jens KoberAndreas WilhelmErhan ÖztopJan PetersPublished in: Auton. Robots (2012)
Keyphrases
- reinforcement learning
- function approximation
- state space
- learning algorithm
- learning process
- optimal policy
- reinforcement learning algorithms
- robotic control
- continuous state
- case based reasoning
- direct policy search
- data sets
- reinforcement learning methods
- control problems
- action selection
- model free
- optimal control
- transfer learning
- sufficient conditions
- multi agent
- similarity measure
- decision trees
- machine learning