Optimized look-ahead trees: Extensions to large and continuous action spaces.
Tobias JungDamien ErnstFrancis MaesPublished in: ADPRL (2013)
Keyphrases
- action space
- state space
- markov decision processes
- real valued
- continuous action
- reinforcement learning
- state and action spaces
- stochastic processes
- continuous state spaces
- continuous state
- control policies
- action selection
- skill learning
- single agent
- markov decision problems
- multi agent
- markov decision process
- machine learning
- optimal policy
- cooperative
- computational complexity
- finite state
- random fields
- policy search
- heuristic search