What is Decidable about Partially Observable Markov Decision Processes with omega-Regular Objectives.
Krishnendu ChatterjeeMartin ChmelikMathieu TracolPublished in: CSL (2013)
Keyphrases
- partially observable markov decision processes
- finite state
- dynamical systems
- decision problems
- reinforcement learning
- planning under uncertainty
- belief state
- belief space
- partially observable stochastic games
- continuous state
- dynamic programming
- optimal policy
- markov decision processes
- partial observability
- partially observable domains
- multi agent
- state space
- stochastic domains
- planning problems
- partially observable markov
- approximate solutions
- partially observable
- dec pomdps
- infinite horizon
- partially observable markov decision process
- np hard
- predictive state representations
- decision theoretic
- optimal control
- model checking
- markov chain
- point based value iteration
- special case