Sample-Efficient Reinforcement Learning Is Feasible for Linearly Realizable MDPs with Limited Revisiting.
Gen LiYuxin ChenYuejie ChiYuantao GuYuting WeiPublished in: NeurIPS (2021)
Keyphrases
- reinforcement learning
- markov decision processes
- state space
- learning algorithm
- function approximation
- optimal policy
- model free
- dynamic programming
- learning process
- reinforcement learning algorithms
- markov decision process
- sufficient conditions
- machine learning
- markov decision problems
- probabilistic planning
- approximate dynamic programming
- factored markov decision processes