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 2408.02529cs.AI
收藏

解释强化学习:一种反事实 Shapley 值方法

Explaining Reinforcement Learning: A Counterfactual Shapley Values Approach

School of Engineering Mathematics and Technology, University of Bristol(布里斯托大学工程数学与技术学院);School of Electronic and Information Engineering, Tongji University(同济大学电子与信息工程学院)

Yiwei Shi, Qi Zhang, Kevin McAreavey, Weiru Liu

AI总结:

提出反事实 Shapley 值(CSV)方法,结合反事实分析与 Shapley 值量化状态维度对动作选择的贡献,并在多个 RL 环境中验证其可解释性效果。

摘要:

本文提出了一种新颖的方法——反事实 Shapley 值(Counterfactual Shapley Values,CSV),通过将反事实分析与 Shapley 值相结合,增强强化学习(RL)的可解释性。该方法旨在量化并比较不同状态维度对各种动作选择的贡献。为了更准确地分析这些影响,我们引入了新的特征值函数:“反事实差异特征值”和“平均反事实差异特征值”。这些函数有助于计算 Shapley 值,以评估最优动作与非最优动作之间的贡献差异。在 GridWorld、FrozenLake 和 Taxi 等多个 RL 领域的实验证明了 CSV 方法的有效性。结果表明,该方法不仅提高了复杂 RL 系统的透明度,还能量化不同决策之间的差异。

英文摘要:

This paper introduces a novel approach Counterfactual Shapley Values (CSV), which enhances explainability in reinforcement learning (RL) by integrating counterfactual analysis with Shapley Values. The approach aims to quantify and compare the contributions of different state dimensions to various action choices. To more accurately analyze these impacts, we introduce new characteristic value functions, the ``Counterfactual Difference Characteristic Value" and the ``Average Counterfactual Difference Characteristic Value." These functions help calculate the Shapley values to evaluate the differences in contributions between optimal and non-optimal actions. Experiments across several RL domains, such as GridWorld, FrozenLake, and Taxi, demonstrate the effectiveness of the CSV method. The results show that this method not only improves transparency in complex RL systems but also quantifies the differences across various decisions.

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。

↑