Minimax Sample Complexity for Turn-based Stochastic Game
Abstract
The empirical success of Multi-agent reinforcement learning is encouraging, while few theoretical guarantees have been revealed. In this work, we prove that the plug-in solver approach, probably the most natural reinforcement learning algorithm, achieves minimax sample complexity for turn-based stochastic game (TBSG). Specifically, we plan in an empirical TBSG by utilizing a `simulator' that allows sampling from arbitrary state-action pair. We show that the empirical Nash equilibrium strategy is an approximate Nash equilibrium strategy in the true TBSG and give both problem-dependent and problem-independent bound. We develop absorbing TBSG and reward perturbation techniques to tackle the complex statistical dependence. The key idea is artificially introducing a suboptimality gap in TBSG and then the Nash equilibrium strategy lies in a finite set.
Keywords
Cite
@article{arxiv.2011.14267,
title = {Minimax Sample Complexity for Turn-based Stochastic Game},
author = {Qiwen Cui and Lin F. Yang},
journal= {arXiv preprint arXiv:2011.14267},
year = {2020}
}
Comments
15 pages