Partially observable Markov decision processes with imprecise parameters.
Hideaki ItohKiyohiko NakamuraPublished in: Artif. Intell. (2007)
Keyphrases
- partially observable markov decision processes
- finite state
- gradient ascent
- belief space
- planning under uncertainty
- expectation maximization
- partial observability
- state space
- belief state
- partially observable domains
- image segmentation
- stochastic domains
- infinite horizon
- planning problems
- decision problems
- markov decision processes
- dynamical systems
- optimal policy
- dynamic programming