Singularly perturbed Markov decision processes in discrete time.
R. H. LiuQian ZhangGeorge YinPublished in: CDC (2001)
Keyphrases
- markov decision processes
- finite state
- semi markov decision processes
- optimal policy
- state space
- reinforcement learning
- stationary policies
- average reward
- reachability analysis
- dynamic programming
- policy iteration
- transition matrices
- markov chain
- average cost
- original data
- partially observable
- finite horizon
- planning under uncertainty
- infinite horizon
- factored mdps
- markov decision process
- action sets
- risk sensitive
- decision theoretic planning
- reinforcement learning algorithms
- sufficient conditions
- decision processes
- action space
- partially observable markov decision processes
- heuristic search
- state and action spaces
- discounted reward
- data mining
- real time dynamic programming