arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2405.02201math.OCcs.LG
收藏

通过稳健平均实现的正则化Q学习

Regularized Q-learning through Robust Averaging

University of Konstanz(康斯坦茨大学)

Peter Schmitt-Förster, Tobias Sutter

AI总结:

提出2RA Q-learning算法,通过分布稳健估计器控制估计偏差,在计算成本相当的情况下收敛于最优策略,性能优于现有方法。

摘要:

我们提出了一种新的Q学习变体,称为2RA Q-learning,以有原则的方式解决了现有Q学习方法的一些弱点。其中一个弱点是无法控制的潜在估计偏差,这通常导致性能不佳。我们为最大期望值项提出了一种分布稳健估计器,使我们能够精确控制引入的估计偏差水平。该分布稳健估计器存在闭式解,使得所提出的算法每次迭代的计算成本与Watkins的Q学习相当。对于表格情况,我们证明了2RA Q学习收敛于最优策略,并分析了其渐近均方误差。最后,我们在各种设置下进行了数值实验,证实了我们的理论发现,并表明2RA Q学习通常优于现有方法。

英文摘要:

We propose a new Q-learning variant, called 2RA Q-learning, that addresses some weaknesses of existing Q-learning methods in a principled manner. One such weakness is an underlying estimation bias which cannot be controlled and often results in poor performance. We propose a distributionally robust estimator for the maximum expected value term, which allows us to precisely control the level of estimation bias introduced. The distributionally robust estimator admits a closed-form solution such that the proposed algorithm has a computational cost per iteration comparable to Watkins' Q-learning. For the tabular case, we show that 2RA Q-learning converges to the optimal policy and analyze its asymptotic mean-squared error. Lastly, we conduct numerical experiments for various settings, which corroborate our theoretical findings and indicate that 2RA Q-learning often performs better than existing methods.

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。

↑