Sample-Path Optimal Stationary Policies in Stable Markov Decision Chains with the Average Reward Criterion.
Rolando Cavazos-CadenaRaúl Montes-de-OcaKarel SladkýPublished in: J. Appl. Probab. (2015)
Keyphrases
- average reward
- sample path
- markov decision processes
- optimality criterion
- stationary policies
- average cost
- optimal policy
- policy iteration
- long run
- finite state
- state space
- reinforcement learning
- markov chain
- dynamic programming
- infinite horizon
- reinforcement learning algorithms
- decision problems
- model free
- markov decision process
- finite horizon
- markov decision problems
- reward function
- lot sizing
- action space
- optimal control
- least squares
- optimal solution