How should Learning Classifier Systems cover a state-action space?
Masaya NakataPier Luca LanziTim KovacsWill Neil BrowneKeiki TakadamaPublished in: CEC (2015)
Keyphrases
- learning classifier systems
- state action space
- reinforcement learning
- function approximation
- classifier systems
- genetic algorithm
- state space
- optimal policy
- real valued
- learning algorithm
- evolutionary algorithm
- temporal difference
- model free
- neural network
- reinforcement learning algorithms
- rule sets
- dynamic programming
- expert systems
- reinforcement learning methods
- policy iteration
- markov decision processes
- learning problems
- text classification
- decision trees
- machine learning