arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2209.13489cs.LGcs.AI
收藏

逆强化学习中使用结构模体的奖励学习

Reward Learning using Structural Motifs in Inverse Reinforcement Learning

University of Toronto(多伦多大学);Vector Institute(向量研究所)

Raeid Saqur

摘要:

逆强化学习(\textit{IRL})问题在过去几年中经历了快速的发展,在机器人、认知和健康等领域有着重要的应用。在这项工作中,我们探讨了目前的IRL方法在学习代理的奖励功能,从专家轨迹描绘长期的,复杂的顺序任务的无效性。我们假设,灌输IRL模型与结构图案捕捉底层任务可以使和提高他们的表现。随后,我们提出了一种新的IRL方法,SMIRL,首先学习的(近似)结构的任务作为一个有限状态自动机(FSA),然后使用的结构基序来解决IRL问题。我们测试我们的模型在离散网格世界和高维连续域环境。我们的经验表明,我们提出的方法成功地学习了所有四个复杂的任务,其中两个基本的IRL基线失败。我们的模型在一个简单的玩具任务上的样本效率也优于基线。我们进一步展示了有前途的测试结果,在一个修改后的连续域的任务与组合奖励函数。

英文摘要:

The Inverse Reinforcement Learning (\textit{IRL}) problem has seen rapid evolution in the past few years, with important applications in domains like robotics, cognition, and health. In this work, we explore the inefficacy of current IRL methods in learning an agent's reward function from expert trajectories depicting long-horizon, complex sequential tasks. We hypothesize that imbuing IRL models with structural motifs capturing underlying tasks can enable and enhance their performance. Subsequently, we propose a novel IRL method, SMIRL, that first learns the (approximate) structure of a task as a finite-state-automaton (FSA), then uses the structural motif to solve the IRL problem. We test our model on both discrete grid world and high-dimensional continuous domain environments. We empirically show that our proposed approach successfully learns all four complex tasks, where two foundational IRL baselines fail. Our model also outperforms the baselines in sample efficiency on a simpler toy task. We further show promising test results in a modified continuous domain on tasks with compositional reward functions.

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。

↑