Using suitable action selection rule in reinforcement learning.
Masayuki OhtaYoichiro KumadaItsuki NodaPublished in: SMC (2003)
Keyphrases
- action selection
- reinforcement learning
- temporal difference
- basal ganglia
- robot soccer
- continuous state and action spaces
- human robot
- state space
- function approximation
- action space
- decision making
- dynamic programming
- action selection mechanism
- reinforcement learning algorithms
- model free
- function approximators
- production rules
- rule sets
- rule learning
- data mining
- markov decision processes
- dynamic environments
- learning algorithm
- machine learning