English

Single-Timescale Stochastic Nonconvex-Concave Optimization for Smooth Nonlinear TD Learning

Machine Learning 2020-08-25 v1 Optimization and Control Machine Learning

Abstract

Temporal-Difference (TD) learning with nonlinear smooth function approximation for policy evaluation has achieved great success in modern reinforcement learning. It is shown that such a problem can be reformulated as a stochastic nonconvex-strongly-concave optimization problem, which is challenging as naive stochastic gradient descent-ascent algorithm suffers from slow convergence. Existing approaches for this problem are based on two-timescale or double-loop stochastic gradient algorithms, which may also require sampling large-batch data. However, in practice, a single-timescale single-loop stochastic algorithm is preferred due to its simplicity and also because its step-size is easier to tune. In this paper, we propose two single-timescale single-loop algorithms which require only one data point each step. Our first algorithm implements momentum updates on both primal and dual variables achieving an O(ε4)O(\varepsilon^{-4}) sample complexity, which shows the important role of momentum in obtaining a single-timescale algorithm. Our second algorithm improves upon the first one by applying variance reduction on top of momentum, which matches the best known O(ε3)O(\varepsilon^{-3}) sample complexity in existing works. Furthermore, our variance-reduction algorithm does not require a large-batch checkpoint. Moreover, our theoretical results for both algorithms are expressed in a tighter form of simultaneous primal and dual side convergence.

Keywords

Cite

@article{arxiv.2008.10103,
  title  = {Single-Timescale Stochastic Nonconvex-Concave Optimization for Smooth Nonlinear TD Learning},
  author = {Shuang Qiu and Zhuoran Yang and Xiaohan Wei and Jieping Ye and Zhaoran Wang},
  journal= {arXiv preprint arXiv:2008.10103},
  year   = {2020}
}

Comments

45 pages; initial draft submitted in Feb, 2020

R2 v1 2026-06-23T18:02:57.978Z