Counterfactual Explanation Trees: Transparent and Consistent Actionable Recourse with Decision Trees
Counterfactual Explanation Trees: Transparent and Consistent Actionable Recourse with Decision Trees
复制标题
DOI：
--
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Kentaro Kanamori;Takuya Takagi;Ken Kobayashi;Yuichi Ike
中科院分区：
文献类型：
--
作者：
Kentaro Kanamori;Takuya Takagi;Ken Kobayashi;Yuichi Ike
反事实解释（CE）是一种事后解释方法，它提供了一个扰动来改变分类器的预测结果。个体可以将扰动解释为获得期望的决策结果的“动作”。现有的CE方法专注于提供一个动作，它是针对给定的单个实例进行优化的。然而，这些CE方法不能解决我们必须同时将动作分配给多个实例的情况。在这种情况下，我们需要一个CE框架，以透明和一致的方式将操作分配给多个实例。在这项研究中，我们提出了反事实解释树（CET），分配有效的行动与决策树。由于决策树的性质，我们的CET有两个优点：（1）透明性：分配动作的原因总结在一个可解释的结构，（2）一致性：这些原因不相互冲突。我们通过两个步骤学习CET：（i）计算多个实例的一个有效动作，（ii）划分实例以平衡有效性和可解释性。数值实验和用户研究表明，我们的CET与现有的方法相比，有效性。
Counterfactual Explanation (CE) is a post-hoc explanation method that provides a perturbation for altering the prediction result of a classifier. An individual can interpret the perturbation as an “action” to obtain the desired decision results. Existing CE methods focus on providing an action, which is optimized for a given single instance. However, these CE methods do not address the case where we have to assign actions to multiple instances simultaneously. In such a case, we need a framework of CE that assigns actions to multiple instances in a transparent and consistent way. In this study, we propose Counterfactual Explanation Tree (CET) that assigns effective actions with decision trees. Due to the properties of decision trees, our CET has two advantages: (1) Transparency: the reasons for assigning actions are summarized in an interpretable structure, and (2) Consistency: these reasons do not con-flict with each other. We learn a CET in two steps: (i) compute one effective action for multiple instances and (ii) partition the instances to balance the effectiveness and inter-pretability. Numerical experiments and user studies demonstrated the efficacy of our CET in comparison with existing methods.