Stochastic optimization of controlled partially observable Markov decision processes.
Peter L. BartlettJonathan BaxterPublished in: CDC (2000)
Keyphrases
- stochastic optimization
- partially observable markov decision processes
- multistage
- finite state
- optimal policy
- dynamical systems
- planning under uncertainty
- belief state
- dynamic programming
- belief space
- decision problems
- reinforcement learning
- partially observable stochastic games
- state space
- stochastic domains
- partially observable domains
- planning problems
- markov decision processes
- point based value iteration
- partially observable
- partially observable markov
- markov chain
- multi agent
- robust optimization
- infinite horizon
- predictive state representations
- dynamic environments
- special case