Off-Policy Confidence Interval Estimation with Confounded Markov Decision Process.
Chengchun ShiJin ZhuYe ShenShikai LuoHongtu ZhuRui SongPublished in: CoRR (2022)
Keyphrases
- markov decision process
- interval estimation
- markov decision processes
- state space
- optimal policy
- infinite horizon
- transition matrices
- reinforcement learning
- finite horizon
- dynamic programming
- collaborative filtering
- temporal difference learning
- policy iteration
- finite state
- partially observable
- reinforcement learning algorithms
- average cost
- initial state
- reward function
- machine learning
- dynamical systems
- sufficient conditions
- decision problems