arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.00614cs.LGcs.AIstat.ML
收藏

基于信息瓶颈的深度强化学习动力学泛化

Dynamics Generalization via Information Bottleneck in Deep Reinforcement Learning

UC Berkeley(加州大学伯克利分校)

Xingyu Lu, Kimin Lee, Pieter Abbeel, Stas Tiomkin

摘要:

尽管深度强化学习(RL)在解决顺序决策问题方面取得了重大进展,但RL代理通常过度适应训练环境,并难以适应新的、不可见的环境。这阻止了RL在现实世界中的鲁棒应用,其中系统动态可能会严重偏离训练设置。在这项工作中,我们的主要贡献是提出了一个信息论正则化目标和一个基于退火的优化方法,以实现更好的推广能力的RL代理。我们展示了我们的方法在不同领域的极端推广效益,从迷宫导航到机器人任务;我们第一次表明,代理可以推广到测试参数超过10个标准差远离训练参数分布。这项工作提供了一种原则性的方法,通过逐渐去除对任务解决来说冗余的信息来提高RL中的泛化能力;它为从训练到极端不同的测试设置的泛化系统研究打开了大门,重点是信息论和机器学习之间的联系。

英文摘要:

Despite the significant progress of deep reinforcement learning (RL) in solving sequential decision making problems, RL agents often overfit to training environments and struggle to adapt to new, unseen environments. This prevents robust applications of RL in real world situations, where system dynamics may deviate wildly from the training settings. In this work, our primary contribution is to propose an information theoretic regularization objective and an annealing-based optimization method to achieve better generalization ability in RL agents. We demonstrate the extreme generalization benefits of our approach in different domains ranging from maze navigation to robotic tasks; for the first time, we show that agents can generalize to test parameters more than 10 standard deviations away from the training parameter distribution. This work provides a principled way to improve generalization in RL by gradually removing information that is redundant for task-solving; it opens doors for the systematic study of generalization from training to extremely different testing settings, focusing on the established connections between information theory and machine learning.

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。

↑