arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1909.09369cs.LGstat.ML
收藏

FACE:可行且可操作的反事实解释

FACE: Feasible and Actionable Counterfactual Explanations

University of Bristol(布里斯托尔大学);University of Ghent(根特大学)

Rafael Poyiadzi, Kacper Sokol, Raul Santos-Rodriguez, Tijl De Bie, Peter Flach

摘要:

反事实解释的工作往往集中在“最接近的可能世界”的原则上,该原则确定了导致预期结果的微小变化。在本文中,我们认为,虽然这种方法最初似乎直观地吸引人,它表现出的缺点,在目前的文献中没有解决。首先,由最先进的系统生成的反事实示例不一定代表底层数据分布,并且因此可能规定无法实现的目标(例如,严重残疾的人寿保险申请人如未能成功投保,可获劝喻多做运动。其次,反事实可能不是基于主体的当前状态和建议状态之间的“可行路径”,使得可诉追索不可行(例如,低技能的不成功的抵押贷款申请人可能被告知要加倍他们的工资,如果不首先提高他们的技能水平,这可能是困难的)。这两个缺点可能会使反事实的解释不切实际,有时甚至是彻头彻尾的冒犯。为了解决这两个主要缺陷,首先,我们提出了一个新的反事实推理研究路线,旨在提供可操作和可行的路径,将选定的实例转换为满足特定目标的实例。第二,我们建议:一种基于密度加权度量定义的最短路径距离来揭示这些“可行路径”的算法上合理的方法。我们的方法生成与底层数据分布一致的反事实,并由变化的“可行路径”支持,这些路径是可以实现的,可以根据手头的问题进行定制。

英文摘要:

Work in Counterfactual Explanations tends to focus on the principle of "the closest possible world" that identifies small changes leading to the desired outcome. In this paper we argue that while this approach might initially seem intuitively appealing it exhibits shortcomings not addressed in the current literature. First, a counterfactual example generated by the state-of-the-art systems is not necessarily representative of the underlying data distribution, and may therefore prescribe unachievable goals(e.g., an unsuccessful life insurance applicant with severe disability may be advised to do more sports). Secondly, the counterfactuals may not be based on a "feasible path" between the current state of the subject and the suggested one, making actionable recourse infeasible (e.g., low-skilled unsuccessful mortgage applicants may be told to double their salary, which may be hard without first increasing their skill level). These two shortcomings may render counterfactual explanations impractical and sometimes outright offensive. To address these two major flaws, first of all, we propose a new line of Counterfactual Explanations research aimed at providing actionable and feasible paths to transform a selected instance into one that meets a certain goal. Secondly, we propose FACE: an algorithmically sound way of uncovering these "feasible paths" based on the shortest path distances defined via density-weighted metrics. Our approach generates counterfactuals that are coherent with the underlying data distribution and supported by the "feasible paths" of change, which are achievable and can be tailored to the problem at hand.

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。

↑