arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 1703.02702cs.LGcs.AIcs.MAcs.RO
收藏

鲁棒的对抗性强化学习

Robust Adversarial Reinforcement Learning

Carnegie Mellon University(卡内基梅隆大学);Google Brain(谷歌大脑);Google Research(谷歌研究院)

Lerrel Pinto, James Davidson, Rahul Sukthankar, Abhinav Gupta

摘要:

深度神经网络与快速模拟和改进的计算相结合,最近在强化学习(RL)领域取得了成功。然而,大多数当前基于RL的方法无法推广,因为:(a)仿真和真实世界之间的差距如此之大,以至于策略学习方法无法转移;(b)即使策略学习在真实世界中完成,数据稀缺也导致从训练到测试场景的推广失败(例如,由于不同的摩擦或物体质量)。受H ∞控制方法的启发,我们注意到,建模误差以及训练和测试场景中的差异都可以被视为系统中的额外力/干扰。本文提出了鲁棒对抗强化学习(RARL)的思想,在那里我们训练一个代理操作的存在下,不稳定的对手,施加干扰力的系统。联合训练的对手得到了加强-也就是说,它学会了一个最佳的不稳定政策。我们制定的策略学习作为一个零和,极大极小的目标函数。在多个环境(InvertedPendulum,HalfCheetah,Swimmer,Hopper和Walker 2d)中进行的大量实验最终证明,我们的方法(a)提高了训练稳定性;(b)对训练/测试条件的差异具有鲁棒性;以及c)即使在没有对手的情况下也优于基线。

英文摘要:

Deep neural networks coupled with fast simulation and improved computation have led to recent successes in the field of reinforcement learning (RL). However, most current RL-based approaches fail to generalize since: (a) the gap between simulation and real world is so large that policy-learning approaches fail to transfer; (b) even if policy learning is done in real world, the data scarcity leads to failed generalization from training to test scenarios (e.g., due to different friction or object masses). Inspired from H-infinity control methods, we note that both modeling errors and differences in training and test scenarios can be viewed as extra forces/disturbances in the system. This paper proposes the idea of robust adversarial reinforcement learning (RARL), where we train an agent to operate in the presence of a destabilizing adversary that applies disturbance forces to the system. The jointly trained adversary is reinforced -- that is, it learns an optimal destabilization policy. We formulate the policy learning as a zero-sum, minimax objective function. Extensive experiments in multiple environments (InvertedPendulum, HalfCheetah, Swimmer, Hopper and Walker2d) conclusively demonstrate that our method (a) improves training stability; (b) is robust to differences in training/test conditions; and c) outperform the baseline even in the absence of the adversary.

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。

↑