A Novel Reinforcement Learning Architecture for Continuous State and Action Spaces.
Víctor Uc-CetinaPublished in: Adv. Artif. Intell. (2013)
Keyphrases
- continuous state and action spaces
- reinforcement learning
- continuous state
- function approximators
- action selection
- reinforcement learning algorithms
- learning capabilities
- state space
- function approximation
- model free
- multi agent
- learning algorithm
- sufficient conditions
- optimal policy
- heuristic search
- temporal difference learning
- markov decision problems
- sparse approximation