Verifiable Planning in Expected Reward Multichain MDPs.
George K. AtiaAndre BeckusIsmail AlkhouriAlvaro VelasquezPublished in: CoRR (2020)
Keyphrases
- markov decision processes
- expected reward
- optimal policy
- finite horizon
- planning under uncertainty
- average reward
- partially observable
- partially observable markov decision processes
- state space
- finite state
- policy iteration
- dynamic programming
- reinforcement learning
- planning problems
- sufficient conditions
- markov decision problems
- average cost
- heuristic search
- long run
- markov decision process
- infinite horizon
- linear program
- initial state
- domain independent
- action space
- decision processes
- special case
- machine learning