A Novel Q-Learning Approach with Continuous States and Actions.
Yi ZhouMeng Joo ErPublished in: CCA (2007)
Keyphrases
- state action
- state action space
- reinforcement learning
- action space
- function approximation
- action selection
- perceptual aliasing
- state space
- state transitions
- initial state
- evaluation function
- continuous state spaces
- cooperative
- action sequences
- internal states
- continuous action
- continuous state and action spaces
- markov decision processes
- continuous state
- multi agent
- goal state
- belief state
- optimal policy
- partial knowledge
- state transition
- stochastic games
- real valued
- stochastic approximation
- temporal difference learning
- reinforcement learning algorithms
- neural network
- model free
- plan recognition
- decision theoretic
- situation calculus
- radial basis function
- function approximators
- markov decision process
- reward function
- markov decision problems
- multiple agents
- potential field
- decision making
- machine learning