arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.08875cs.LGcs.AIstat.ML
收藏

基于模型的对抗性Meta-强化学习

Model-based Adversarial Meta-Reinforcement Learning

Tsinghua University(清华大学);Stanford University(斯坦福大学)

Zichuan Lin, Garrett Thomas, Guangwen Yang, Tengyu Ma

摘要:

Meta-强化学习(meta-RL)旨在从多个训练任务中学习有效适应未知测试任务的能力。尽管取得了成功,但已知现有的元RL算法对任务分布偏移敏感。当测试任务分布与训练任务分布不同时,性能可能会显著下降。为了解决这个问题,本文提出了基于模型的对抗性Meta-强化学习(AdMRL),我们的目标是最大限度地减少最坏情况下的次优差距-最佳回报与算法适应后实现的回报之间的差异-在一个任务家族中的所有任务中,使用基于模型的方法。我们提出了一个极大极小目标,并通过在固定任务上学习动态模型和为当前模型找到对抗性任务之间进行交替来优化它-该任务的模型诱导的策略是最大次优的。假设家庭的任务是参数化的,我们推导出一个公式的梯度的次优性相对于任务参数通过隐函数定理,并显示如何梯度估计可以有效地实现共轭梯度法和一种新的使用的REINFORCE估计。我们在几个连续控制基准上评估了我们的方法,并在所有任务的最坏情况下的性能,分布任务的泛化能力,以及训练和测试时间的样本效率方面证明了它的有效性,超过了现有的最先进的元RL算法。

英文摘要:

Meta-reinforcement learning (meta-RL) aims to learn from multiple training tasks the ability to adapt efficiently to unseen test tasks. Despite the success, existing meta-RL algorithms are known to be sensitive to the task distribution shift. When the test task distribution is different from the training task distribution, the performance may degrade significantly. To address this issue, this paper proposes Model-based Adversarial Meta-Reinforcement Learning (AdMRL), where we aim to minimize the worst-case sub-optimality gap -- the difference between the optimal return and the return that the algorithm achieves after adaptation -- across all tasks in a family of tasks, with a model-based approach. We propose a minimax objective and optimize it by alternating between learning the dynamics model on a fixed task and finding the adversarial task for the current model -- the task for which the policy induced by the model is maximally suboptimal. Assuming the family of tasks is parameterized, we derive a formula for the gradient of the suboptimality with respect to the task parameters via the implicit function theorem, and show how the gradient estimator can be efficiently implemented by the conjugate gradient method and a novel use of the REINFORCE estimator. We evaluate our approach on several continuous control benchmarks and demonstrate its efficacy in the worst-case performance over all tasks, the generalization power to out-of-distribution tasks, and in training and test time sample efficiency, over existing state-of-the-art meta-RL algorithms.

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。

↑