Solving Discounted Stochastic Two-Player Games with Near-Optimal Time and Sample Complexity.
Aaron SidfordMengdi WangLin F. YangYinyu YePublished in: CoRR (2019)
Keyphrases
- sample complexity
- two player games
- learning problems
- theoretical analysis
- special case
- upper bound
- learning algorithm
- supervised learning
- generalization error
- lower bound
- active learning
- evaluation function
- dynamic programming
- sample size
- training data
- learning tasks
- optimal policy
- machine learning algorithms
- training examples
- markov decision processes
- feature space