Improving reinforcement learning algorithms by the use of data mining techniques for feature and action selection.
Davi C. de L. VieiraPaulo J. L. AdeodatoPaulo M. GoncalvesPublished in: SMC (2010)
Keyphrases
- action selection
- reinforcement learning algorithms
- temporal difference
- reinforcement learning
- data mining techniques
- total reward
- basal ganglia
- robot soccer
- data mining
- function approximation
- reinforcement learning problems
- state space
- model free
- eligibility traces
- markov decision processes
- reinforcement learning methods
- policy iteration
- action selection mechanism
- policy search
- learning algorithm
- action space
- association rules
- reward function
- dynamic programming
- neural network
- multi agent
- feature vectors
- optimal policy