On strong average optimality of markov decision processes with unbounded costs.
Mrinal K. GhoshSteven I. MarcusPublished in: Oper. Res. Lett. (1992)
Keyphrases
- average cost
- markov decision processes
- finite state
- state space
- optimal policy
- finite horizon
- dynamic programming
- infinite horizon
- policy iteration
- reinforcement learning
- average reward
- reachability analysis
- risk sensitive
- planning under uncertainty
- initial state
- reinforcement learning algorithms
- total cost
- transition matrices
- model based reinforcement learning
- long run
- markov decision problems
- decision processes
- stationary policies
- multistage
- decision theoretic planning
- factored mdps
- action sets
- partially observable
- optimal control
- linear program
- markov decision process
- partially observable markov decision processes
- discounted reward
- linear programming
- state abstraction
- state and action spaces
- reward function