arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1702.03006cs.LG
收藏

无重要性采样比的多步离策略学习

Multi-step Off-policy Learning Without Importance Sampling Ratios

University of Alberta(阿尔伯塔大学)

Ashique Rupam Mahmood, Huizhen Yu, Richard S. Sutton

摘要:

为了从探索性数据中估计策略的价值函数,大多数无模型的非策略算法依赖于重要性采样,其中使用重要性采样比率通常导致具有严重方差的估计。因此,期望在不使用比率的情况下学习偏离策略。然而,这样的算法不存在用于具有函数逼近的多步学习。在本文中,我们介绍了第一个这样的算法基于时间差(TD)学习更新。我们表明,明确使用的重要性采样比可以消除通过改变TD更新中的自举量的动作依赖的方式。我们的新算法实现稳定使用两个时间尺度的梯度为基础的TD更新。一个先前的算法的基础上查找表表示称为树备份也可以检索使用动作相关的引导,成为我们的算法的一个特殊情况。在两个具有挑战性的非策略任务中,我们证明了我们的算法是稳定的,有效地避免了大方差问题,并且可以比最先进的算法表现得更好。

英文摘要:

To estimate the value functions of policies from exploratory data, most model-free off-policy algorithms rely on importance sampling, where the use of importance sampling ratios often leads to estimates with severe variance. It is thus desirable to learn off-policy without using the ratios. However, such an algorithm does not exist for multi-step learning with function approximation. In this paper, we introduce the first such algorithm based on temporal-difference (TD) learning updates. We show that an explicit use of importance sampling ratios can be eliminated by varying the amount of bootstrapping in TD updates in an action-dependent manner. Our new algorithm achieves stability using a two-timescale gradient-based TD update. A prior algorithm based on lookup table representation called Tree Backup can also be retrieved using action-dependent bootstrapping, becoming a special case of our algorithm. In two challenging off-policy tasks, we demonstrate that our algorithm is stable, effectively avoids the large variance issue, and can perform substantially better than its state-of-the-art counterpart.

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。

↑