arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.02863cs.LGstat.ML
收藏

具有显式策略估计的逆强化学习

Inverse Reinforcement Learning with Explicit Policy Estimates

Carnegie Mellon University(卡内基梅隆大学);Sony Corporation(索尼公司)

Navyata Sanghvi, Shinnosuke Usami, Mohit Sharma, Joachim Groeger, Kris Kitani

摘要:

在机器学习和经济学中,已经独立地开发了用于解决逆强化学习(IRL)问题的各种方法。特别是,最大因果熵IRL方法是基于熵最大化的观点,而经济学领域的相关进展则假设存在未观察到的行动冲击来解释专家行为(嵌套不动点算法,条件选择概率方法,嵌套伪Likestive算法)。在这项工作中,我们在这两个领域的这些相关方法之间建立了以前未知的联系。我们实现这一点,表明它们都属于一类优化问题,其特征在于一个共同的形式的目标,相关的政策和目标梯度。我们证明了关键的计算和算法之间的差异,由于最佳软价值函数的近似的方法,并描述了如何导致更有效的算法。使用的见解,从我们的研究这一类的优化问题,我们确定了各种问题的情况下,并调查每种方法的适用性,这些问题。

英文摘要:

Various methods for solving the inverse reinforcement learning (IRL) problem have been developed independently in machine learning and economics. In particular, the method of Maximum Causal Entropy IRL is based on the perspective of entropy maximization, while related advances in the field of economics instead assume the existence of unobserved action shocks to explain expert behavior (Nested Fixed Point Algorithm, Conditional Choice Probability method, Nested Pseudo-Likelihood Algorithm). In this work, we make previously unknown connections between these related methods from both fields. We achieve this by showing that they all belong to a class of optimization problems, characterized by a common form of the objective, the associated policy and the objective gradient. We demonstrate key computational and algorithmic differences which arise between the methods due to an approximation of the optimal soft value function, and describe how this leads to more efficient algorithms. Using insights which emerge from our study of this class of optimization problems, we identify various problem scenarios and investigate each method's suitability for these problems.

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。

↑