arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2210.04820cs.LG
收藏

用于减少复杂问题中深度强化学习方差的长期N步代理阶段奖励

Long N-step Surrogate Stage Reward to Reduce Variances of Deep Reinforcement Learning in Complex Problems

Arizona State University(亚利桑那州立大学)

Junmin Zhong, Ruofan Wu, Jennie Si

AI总结:

本文提出一种长期N步代理阶段(LNSS)奖励方法,通过有效应对复杂环境动态来减少深度强化学习中的方差,在多个复杂基准环境中提升了总奖励、收敛速度和稳定性,并从理论上证明了其能指数级降低Q值方差上界。

摘要:

强化学习中的高方差已被证明会阻碍成功收敛并损害任务性能。由于奖励信号在学习行为中起着重要作用,多步方法已被考虑用于缓解该问题,并认为比单步方法更有效。然而,针对这一重要方面缺乏全面系统的研究,以证明多步方法在解决高度复杂连续控制问题上的有效性。在本研究中,我们引入了一种新的长期N步代理阶段(LNSS)奖励方法,以有效应对复杂的环境动态,而以往的方法通常仅适用于有限步数。LNSS方法简单、计算成本低,适用于基于价值或策略梯度的强化学习。我们在OpenAI Gym和DeepMind Control Suite中系统评估了LNSS,以解决一些复杂的基准环境,这些环境通常难以通过深度强化学习(DRL)获得良好结果。我们证明了LNSS在总奖励、收敛速度和变异系数(CV)方面的性能提升。我们还提供了分析见解,说明LNSS如何从相应的单步方法中指数级地减少Q值方差的上界。

英文摘要:

High variances in reinforcement learning have shown impeding successful convergence and hurting task performance. As reward signal plays an important role in learning behavior, multi-step methods have been considered to mitigate the problem, and are believed to be more effective than single step methods. However, there is a lack of comprehensive and systematic study on this important aspect to demonstrate the effectiveness of multi-step methods in solving highly complex continuous control problems. In this study, we introduce a new long $N$-step surrogate stage (LNSS) reward approach to effectively account for complex environment dynamics while previous methods are usually feasible for limited number of steps. The LNSS method is simple, low computational cost, and applicable to value based or policy gradient reinforcement learning. We systematically evaluate LNSS in OpenAI Gym and DeepMind Control Suite to address some complex benchmark environments that have been challenging to obtain good results by DRL in general. We demonstrate performance improvement in terms of total reward, convergence speed, and coefficient of variation (CV) by LNSS. We also provide analytical insights on how LNSS exponentially reduces the upper bound on the variances of Q value from a respective single step method

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。

↑