Settling the Horizon-Dependence of Sample Complexity in Reinforcement Learning.
Yuanzhi LiRuosong WangLin F. YangPublished in: CoRR (2021)
Keyphrases
- sample complexity
- reinforcement learning
- learning problems
- learning algorithm
- supervised learning
- theoretical analysis
- pac learning
- vc dimension
- special case
- lower bound
- generalization error
- upper bound
- sequential decision problems
- active learning
- training examples
- state space
- optimal policy
- concept classes
- machine learning
- pac learnability
- dynamic programming
- irrelevant features
- multi agent
- training data
- learning tasks
- markov decision processes
- machine learning algorithms
- learning process
- small number
- kernel methods
- sample size
- sample complexity bounds
- number of irrelevant features
- semi supervised learning