Finding the Near Optimal Policy via Adaptive Reduced Regularization in MDPs.
Wenhao YangXiang LiGuangzeng XieZhihua ZhangPublished in: CoRR (2020)
Keyphrases
- optimal policy
- markov decision processes
- finite horizon
- state space
- reinforcement learning
- decision problems
- markov decision process
- infinite horizon
- markov decision problems
- average reward
- finite state
- long run
- dynamic programming
- state dependent
- average cost
- sufficient conditions
- bayesian reinforcement learning
- policy iteration
- multistage
- initial state
- dynamic programming algorithms
- reward function
- discount factor
- serial inventory systems
- action space
- partially observable
- function approximation
- markov chain