On polynomial cases of the unichain classification problem for Markov Decision Processes.
Eugene A. FeinbergFenghsu YangPublished in: Oper. Res. Lett. (2008)
Keyphrases
- markov decision processes
- finite state
- average cost
- optimal policy
- reinforcement learning
- state space
- dynamic programming
- initial state
- decision processes
- policy iteration
- finite horizon
- markov chain
- partially observable
- model based reinforcement learning
- factored mdps
- transition matrices
- action sets
- stationary policies
- machine learning
- long run
- infinite horizon
- average reward
- planning under uncertainty
- decision theoretic planning
- risk sensitive
- state and action spaces
- reachability analysis
- optimal control
- state abstraction
- reinforcement learning algorithms