分离特征重要性中的独立贡献、交互作用与依赖关系，提升解释可信度。
Disentangling Interactions and Dependencies in Feature Attribution
- 提出DIP分解法，将特征重要性拆解为独立、交互和依赖三部分。
- 证明该分解唯一且可实际估算，避免传统方法的混淆误差。
- 新可视化方式清晰展示各类贡献，适合模型可解释性研究者使用。
在可解释机器学习中，全局特征重要性方法试图确定每个特征对目标变量预测的贡献，从而为每个特征生成一个重要性得分。但预测目标变量通常需要多个特征之间的相互作用（如XOR函数），且特征间可能存在复杂统计依赖，使得某些特征可部分替代另一些特征。在常用的特征重要性得分中，这些协同效应与特征的独立贡献混杂在一起，容易导致误读。本文推导出DIP——一种新的个体特征重要性得分的数学分解方法，可分离出三个组成部分：独立贡献、交互贡献与依赖贡献。我们证明了DIP分解的唯一性，并展示了其在实践中如何估计。基于此结果，我们提出了一个新的特征重要性可视化方法，能清晰呈现各类贡献。
原文摘要 · Abstract (English)
In explainable machine learning, global feature importance methods try to determine how much each individual feature contributes to predicting the target variable, resulting in one importance score for each feature. But often, predicting the target variable requires interactions between several features (such as in the XOR function), and features might have complex statistical dependencies that allow to partially replace one feature with another one. In commonly used feature importance scores these cooperative effects are conflated with the features' individual contributions, making them prone to misinterpretations. In this work, we derive DIP, a new mathematical decomposition of individual feature importance scores that disentangles three components: the standalone contribution and the contributions stemming from interactions and dependencies. We prove that the DIP decomposition is unique and show how it can be estimated in practice. Based on these results, we propose a new visualization of feature importance scores that clearly illustrates the different contributions.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。