Hybrid Value Estimation for Off-policy Evaluation and Offline Reinforcement Learning.
Xue-Kun JinXu-Hui LiuShengyi JiangYang YuPublished in: CoRR (2022)
Keyphrases
- policy evaluation
- reinforcement learning
- semi parametric
- temporal difference
- least squares
- model free
- monte carlo
- function approximation
- markov decision processes
- policy iteration
- td learning
- variance reduction
- optimal policy
- state space
- density estimation
- constrained optimization
- partially observable markov decision processes
- multi agent
- regression model
- importance sampling
- reinforcement learning algorithms
- statistical inference
- function approximators
- step size
- finite state
- optimal control
- machine learning
- evaluation function
- transfer learning
- parameter estimation
- search space
- learning algorithm