Reinforcement Learning Using Approximate Belief States.
Andres C. RodriguezRonald ParrDaphne KollerPublished in: NIPS (1999)
Keyphrases
- belief state
- reinforcement learning
- state space
- partial observability
- partially observable
- point based value iteration
- partially observable markov decision processes
- continuous state
- belief space
- reinforcement learning algorithms
- state action
- belief revision
- markov decision processes
- partial knowledge
- partially observable markov decision process
- function approximation
- dynamic programming
- markov decision process
- optimal policy
- model free
- learning algorithm
- multi agent
- dynamical systems
- optimal control
- initial state
- action space
- markov chain