Extreme Occupation Measures in Markov Decision Processes with an Absorbing State.
Alexei B. PiunovskiyYi ZhangPublished in: SIAM J. Control. Optim. (2024)
Keyphrases
- markov decision processes
- state space
- action space
- optimal policy
- dynamic programming
- finite state
- reinforcement learning
- partially observable
- policy iteration
- markov decision process
- state abstraction
- real time dynamic programming
- planning under uncertainty
- transition matrices
- markov chain
- factored mdps
- total reward
- average reward
- decision theoretic planning
- reachability analysis
- risk sensitive
- discounted reward
- infinite horizon
- reinforcement learning algorithms
- heuristic search
- finite horizon
- decision problems
- decision processes
- belief state
- average cost
- action sets