arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2606.05967stat.MLcs.LG
收藏

具有线性函数逼近、通用学习步长和独立同分布样本的TD(0)的快速鲁棒收敛速率

Fast and Robust Convergence Rate for TD(0) with Linear Function Approximation, Universal Learning Steps and I.I.D. Samples

L2S, INRIA U2IS, ENSTA Université Paris-Saclay, CentraleSupelec, Institut Polytechnique de Paris(L2S、INRIA U2IS、ENSTA巴黎萨克雷大学、中央苏格兰、巴黎理工学院)

Ziad Kobeissi, Éloïse Berthier

AI总结:

针对线性函数逼近的TD(0)算法,在独立同分布样本和常数学习步长下,提出一种均方误差的快速(1/k阶)、鲁棒(不依赖最小特征值)且尖锐(乘性常数小于11)的收敛速率,并引入PCTD(0)变体以在强混合假设下获得更好收敛性。

摘要:

本文研究了具有线性函数逼近(LFA)的TD(0)时序差分方法的有限时间行为。我们考虑策略内独立同分布(i.i.d.)样本、常数学习步长和Polyak-Juditsky平均方法。我们为近似函数的均方误差(MSE)建立了一个新的收敛速率,该速率(i)快速,即具有迭代次数k的最优依赖性(即1/k阶),(ii)对病态条件鲁棒:仅依赖于初始误差和模型无关常数,以及(iii)尖锐,乘性常数小于11。特别地,与TD(0)文献中所有现有的O(1/k)速率不同,它不依赖于线性参数化的非中心协方差矩阵的最小特征值。我们还引入了PCTD(0),这是TD(0)的一个变体,在马尔可夫链的强混合附加假设下具有更好的收敛性质。

英文摘要:

In this paper, we study the finite-time behavior of the TD(0) temporal-difference method with linear function approximation (LFA). We consider on-policy independent and identically distributed (i.i.d.) samples, a constant learning step, and the Polyak-Juditsky averaging method. We establish a new convergence rate, for the Mean-Square Error (MSE) on the approximated function, that is (i) fast in the sense that it admits an optimal dependency in the number of iterations k (i.e., of order 1/k), (ii) robust to ill-conditioning: it only depends on an initial error and modelindependent constants and (iii) sharp up to a multiplicative constant lower than 11. In particular, it does not depend on the smallest eigenvalue of the uncentered covariance matrix of the linear parametrization, unlike all pre-existing O(1/k) rates in the TD(0) literature. We also introduce PCTD(0), a variant of TD(0), which benefits from better convergence properties under an additional assumption of strong mixing on the Markov Chain.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑