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 2004.14990cs.LGstat.ML
收藏

基于扩充数据的强化学习

Reinforcement Learning with Augmented Data

UC Berkeley(加州大学伯克利分校);New York University(纽约大学)

Michael Laskin, Kimin Lee, Adam Stooke, Lerrel Pinto, Pieter Abbeel, Aravind Srinivas

摘要:

从视觉观察中学习是强化学习(RL)中的一个基本而又具有挑战性的问题。尽管结合卷积神经网络的算法进步已被证明是成功的秘诀,但目前的方法仍然缺乏两个方面:(a)学习的数据效率和(b)对新环境的泛化。为此,我们提出了增强数据(RAD)的强化学习,这是一个简单的即插即用模块,可以增强大多数RL算法。我们进行了第一次广泛的研究一般数据增强RL基于像素和基于状态的输入,并介绍了两个新的数据增强-随机平移和随机幅度缩放。我们表明,随机平移、裁剪、颜色抖动、补丁剪切、随机卷积和幅度缩放等增强可以使简单的RL算法在常见的基准测试中优于复杂的最先进的方法。RAD在基于像素控制的DeepMind Control Suite基准测试以及基于状态控制的OpenAI Gym基准测试的数据效率和最终性能方面树立了新的先进水平。我们进一步证明了RAD在几个OpenAI ProcGen基准测试中显着提高了现有方法的测试时间泛化能力。我们的RAD模块和培训代码可以在https://www.github.com/MishaLaskin/rad上找到。

英文摘要:

Learning from visual observations is a fundamental yet challenging problem in Reinforcement Learning (RL). Although algorithmic advances combined with convolutional neural networks have proved to be a recipe for success, current methods are still lacking on two fronts: (a) data-efficiency of learning and (b) generalization to new environments. To this end, we present Reinforcement Learning with Augmented Data (RAD), a simple plug-and-play module that can enhance most RL algorithms. We perform the first extensive study of general data augmentations for RL on both pixel-based and state-based inputs, and introduce two new data augmentations - random translate and random amplitude scale. We show that augmentations such as random translate, crop, color jitter, patch cutout, random convolutions, and amplitude scale can enable simple RL algorithms to outperform complex state-of-the-art methods across common benchmarks. RAD sets a new state-of-the-art in terms of data-efficiency and final performance on the DeepMind Control Suite benchmark for pixel-based control as well as OpenAI Gym benchmark for state-based control. We further demonstrate that RAD significantly improves test-time generalization over existing methods on several OpenAI ProcGen benchmarks. Our RAD module and training code are available at https://www.github.com/MishaLaskin/rad.

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。

↑