arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1809.02591cs.LGcs.AIstat.ML
收藏

策略泛化的学习不变性

Learning Invariances for Policy Generalization

Microsoft Research Montreal(微软研究院蒙特利尔分院)

Remi Tachet, Philip Bachman, Harm van Seijen

摘要:

虽然最近的进展催生了非常强大的机器学习系统,但这些代理仍然非常专业,无法将他们获得的知识转移到类似但看不见的任务中。在本文中,我们研究了一个简单的强化学习问题,并专注于学习策略,编码适当的不变性推广到不同的设置。我们评估了三种潜在的策略推广方法:数据增强,元学习和对抗训练。我们发现我们的数据增强方法是有效的,并研究了元学习和对抗学习作为替代任务不可知方法的潜力。

英文摘要:

While recent progress has spawned very powerful machine learning systems, those agents remain extremely specialized and fail to transfer the knowledge they gain to similar yet unseen tasks. In this paper, we study a simple reinforcement learning problem and focus on learning policies that encode the proper invariances for generalization to different settings. We evaluate three potential methods for policy generalization: data augmentation, meta-learning and adversarial training. We find our data augmentation method to be effective, and study the potential of meta-learning and adversarial learning as alternative task-agnostic approaches.

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。

↑