Online least-squares policy iteration for reinforcement learning control.
Lucian BusoniuDamien ErnstBart De SchutterRobert BabuskaPublished in: ACC (2010)
Keyphrases
- reinforcement learning
- model free
- reinforcement learning algorithms
- control problems
- temporal difference
- reinforcement learning methods
- control policy
- optimal control
- policy iteration
- function approximation
- real time
- online learning
- state space
- robot control
- control system
- learning algorithm
- machine learning
- action selection
- markov decision processes
- transfer learning
- optimal policy
- finite sample
- continuous state spaces
- balancing exploration and exploitation