CEGAR for compositional analysis of qualitative properties in Markov decision processes.
Krishnendu ChatterjeeMartin ChmelikPrzemyslaw DacaPublished in: Formal Methods Syst. Des. (2015)
Keyphrases
- markov decision processes
- finite state
- reinforcement learning
- dynamic programming
- optimal policy
- reachability analysis
- decision theoretic planning
- decision processes
- risk sensitive
- planning under uncertainty
- policy iteration
- finite horizon
- markov decision process
- average reward
- infinite horizon
- multistage
- state space
- multi agent