[论文解读] The Meta-Evaluation Problem in Explainable AI: Identifying Reliable Estimators with MetaQuantus

Anna Karin Hedström, Philine Lou Bommer+4 more|ArXiv.org|Feb 14, 2023

Explainable Artificial Intelligence (XAI)被引用 9

一句话总结

该论文为 XAI 估计量定义元评估问题并提出 MetaQuantus，这是一个通过对噪声的鲁棒性和对随机性的反应性来评估估计量可靠性的框架，在没有 ground truth 解释的情况下实现评估。

ABSTRACT

One of the unsolved challenges in the field of Explainable AI (XAI) is determining how to most reliably estimate the quality of an explanation method in the absence of ground truth explanation labels. Resolving this issue is of utmost importance as the evaluation outcomes generated by competing evaluation methods (or ''quality estimators''), which aim at measuring the same property of an explanation method, frequently present conflicting rankings. Such disagreements can be challenging for practitioners to interpret, thereby complicating their ability to select the best-performing explanation method. We address this problem through a meta-evaluation of different quality estimators in XAI, which we define as ''the process of evaluating the evaluation method''. Our novel framework, MetaQuantus, analyses two complementary performance characteristics of a quality estimator: its resilience to noise and reactivity to randomness, thus circumventing the need for ground truth labels. We demonstrate the effectiveness of our framework through a series of experiments, targeting various open questions in XAI such as the selection and hyperparameter optimisation of quality estimators. Our work is released under an open-source license (https://github.com/annahedstroem/MetaQuantus) to serve as a development tool for XAI- and Machine Learning (ML) practitioners to verify and benchmark newly constructed quality estimators in a given explainability context. With this work, we provide the community with clear and theoretically-grounded guidance for identifying reliable evaluation methods, thus facilitating reproducibility in the field.

研究动机与目标

• 由于评估者之间的排名不一致，激发对 XAI 质量估计量进行元评估的需求。
• 提出一个正式框架，在没有 ground-truth 解释的情况下评估估计量的可靠性。
• 引入失败模式和可靠性度量来指导估计量的选择与调优。
• 演示元评估如何在跨任务中支持估计量选择和超参数优化。

提出的方法

• 用可验证空间与不可验证空间的有向无环图（DAG）对归因式解释的评估问题进行建模。
• 为估计量定义两种失败模式：对噪声的鲁棒性（NR）和对对手的反应性（AR）。
• 提出对可验证空间的微扰和破坏性扰动，以对估计量进行压力测试（输入或模型扰动）。
• 引入内部一致性（IAC）和外部一致性（IEC）标准，通过统计检验衡量估计量对扰动的反应。
• 将 NR 与 AR 评估整合成一个单一的 Meta-Consistency（MC）分数，以概括估计量的可靠性。
• 提供实用的扰动方案和评估步骤，包括如何计算 p 值和基于排序的度量。

实验结果

研究问题

• RQ1 在缺乏真实 ground truth 的情况下，我们如何可靠地评估解释方法的质量估计量的可靠性？
• RQ2 哪些失败模式能够最好地捕捉在受控扰动下估计量的鲁棒性和敏感性？
• RQ3 元评估框架是否能在跨 XAI 任务中指导估计量的选择和超参数调优？

主要发现

• 元评估框架通过探测对噪声的鲁棒性和对随机性的反应性，能够识别出可靠的估计量。
• 该框架使用内部一致性和外部一致性度量来量化扰动下的估计量性能。
• 一个单一的 Meta-Consistency 分数将 NR 与 AR 洞见融合，以比较估计量。
• 实验表明，该框架在跨数据集和模型的估计量选择与超参数优化方面具有实用性。