Reinforcement Distribution in Continuous State Action Space Fuzzy Q-Learning: A Novel Approach.
Andrea BonariniFrancesco MontroneMarcello RestelliPublished in: WILF (2005)
Keyphrases
- state action space
- reinforcement learning
- function approximation
- continuous state and action spaces
- fuzzy sets
- action space
- state space
- multi agent
- membership functions
- reinforcement learning algorithms
- actor critic
- learning algorithm
- machine learning
- temporal difference learning
- fuzzy rules
- model free
- supervised learning
- fuzzy logic
- markov decision processes
- learning problems
- optimal control
- optimal policy
- fuzzy systems
- temporal difference
- control problems
- mountain car
- fuzzy controller
- average reward
- single agent
- partially observable
- learning tasks
- linear combination
- artificial neural networks