Policy gradient fuzzy reinforcement learning control of humanoid walking.
Dusko KaticAleksandar D. RodicPublished in: ICONS (2009)
Keyphrases
- policy gradient
- reinforcement learning
- optimal control
- actor critic
- control problems
- reinforcement learning algorithms
- policy search
- function approximation
- humanoid robot
- control system
- reinforcement learning methods
- policy gradient methods
- gradient method
- adaptive control
- control strategies
- model free
- control strategy
- markov decision processes
- control method
- fuzzy logic
- temporal difference
- function approximators
- approximation methods
- variance reduction
- model free reinforcement learning
- dynamic programming
- state space
- fuzzy sets
- single agent
- action selection
- fuzzy rules
- control policy
- neural network
- fuzzy controller
- multi agent
- fuzzy neural network