On the Sample Complexity of Learning Infinite-horizon Discounted Linear Kernel MDPs
Yuanzhou Chen, Jiafan He, Quanquan Gu
Abstract
We study reinforcement learning for infinite-horizon discounted linear kernel MDPs, where the transition probability function is linear in a predefined feature mapping. Existing UCLK \citep{zhou2020provably} algorithm for this setting only has a regret guarantee, which cannot lead to a tight sample complexity bound. In this paper, we extend the uniform-PAC sample complexity from episodic setting to the infinite-horizon discounted setting, and propose a novel algorithm dubbed UPAC-UCLK that achieves an $\Tilde{O}\big(d^2/((1-\gamma)^4\epsilon^2)+1/((1-\gamma)^6\epsilon^2)\big)$ uniform-PAC sample complexity, where $d$ is the dimension of the feature mapping, $\gamma \in(0,1)$ is the discount factor of the MDP and $\epsilon$ is the accuracy parameter. To the best of our knowledge, this is the first $\tilde{O}(1/\epsilon^2)$ sample complexity bound for learning infinite-horizon discounted MDPs with linear function approximation (without access to the generative model).
BibTeX
@InProceedings{pmlr-v162-chen22f,
title = {On the Sample Complexity of Learning Infinite-horizon Discounted Linear Kernel {MDP}s},
author = {Chen, Yuanzhou and He, Jiafan and Gu, Quanquan},
booktitle = {Proceedings of the 39th International Conference on Machine Learning},
pages = {3149--3183},
year = {2022},
editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan},
volume = {162},
series = {Proceedings of Machine Learning Research},
month = {17--23 Jul},
publisher = {PMLR},
pdf = {https://proceedings.mlr.press/v162/chen22f/chen22f.pdf},
url = {https://proceedings.mlr.press/v162/chen22f.html},
abstract = {We study reinforcement learning for infinite-horizon discounted linear kernel MDPs, where the transition probability function is linear in a predefined feature mapping. Existing UCLK \citep{zhou2020provably} algorithm for this setting only has a regret guarantee, which cannot lead to a tight sample complexity bound. In this paper, we extend the uniform-PAC sample complexity from episodic setting to the infinite-horizon discounted setting, and propose a novel algorithm dubbed UPAC-UCLK that achieves an $\Tilde{O}\big(d^2/((1-\gamma)^4\epsilon^2)+1/((1-\gamma)^6\epsilon^2)\big)$ uniform-PAC sample complexity, where $d$ is the dimension of the feature mapping, $\gamma \in(0,1)$ is the discount factor of the MDP and $\epsilon$ is the accuracy parameter. To the best of our knowledge, this is the first $\tilde{O}(1/\epsilon^2)$ sample complexity bound for learning infinite-horizon discounted MDPs with linear function approximation (without access to the generative model).}
}