Exact Decomposition Approaches for Markov Decision Processes: A Survey.
Cherki DaouiMohammed AbbadMohamed TkiouatPublished in: Adv. Oper. Res. (2010)
Keyphrases
- markov decision processes
- finite state
- reinforcement learning
- transition matrices
- optimal policy
- state space
- reinforcement learning algorithms
- decision processes
- policy iteration
- dynamic programming
- planning under uncertainty
- reachability analysis
- finite horizon
- search algorithm
- factored mdps
- decision theoretic planning
- risk sensitive
- average reward
- partially observable
- infinite horizon
- sufficient conditions
- action space
- average cost
- utility function