arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2011.10024cs.LGcs.RO
收藏

Parrot:用于强化学习的数据驱动行为先验

Parrot: Data-Driven Behavioral Priors for Reinforcement Learning

University of California, Berkeley(加州大学伯克利分校)

Avi Singh, Huihan Liu, Gaoyue Zhou, Albert Yu, Nicholas Rhinehart, Sergey Levine

摘要:

强化学习为灵活的决策和控制提供了一个通用框架,但需要为智能体需要学习的每个新任务收集大量的数据汇集。在其他机器学习领域,如自然语言处理或计算机视觉,对先前收集的大型数据集进行预训练以引导新任务的学习已经成为一种强大的范例,以减少学习新任务时的数据需求。在本文中,我们提出了以下问题:我们如何为RL代理启用类似有用的预训练?我们提出了一种用于预训练行为先验的方法,该方法可以从广泛的先前看到的任务中捕获在成功试验中观察到的复杂输入-输出关系,并且我们展示了这种学习的先验如何用于快速学习新任务,而不会阻碍RL代理尝试新行为的能力。我们证明了我们的方法在挑战机器人操作领域的有效性,涉及图像观测和稀疏奖励函数,我们的方法在很大程度上优于以前的作品。

英文摘要:

Reinforcement learning provides a general framework for flexible decision making and control, but requires extensive data collection for each new task that an agent needs to learn. In other machine learning fields, such as natural language processing or computer vision, pre-training on large, previously collected datasets to bootstrap learning for new tasks has emerged as a powerful paradigm to reduce data requirements when learning a new task. In this paper, we ask the following question: how can we enable similarly useful pre-training for RL agents? We propose a method for pre-training behavioral priors that can capture complex input-output relationships observed in successful trials from a wide range of previously seen tasks, and we show how this learned prior can be used for rapidly learning new tasks without impeding the RL agent's ability to try out novel behaviors. We demonstrate the effectiveness of our approach in challenging robotic manipulation domains involving image observations and sparse reward functions, where our method outperforms prior works by a substantial margin.

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。

↑