CID: Measuring Feature Importance Through Counterfactual Distributions

TL;DR

This paper introduces CID, a new post-hoc local feature importance method that uses counterfactual distributions and a rigorous dissimilarity measure to provide more faithful explanations of machine learning models.

Contribution

The paper presents a novel feature importance measure based on counterfactual distributions and a mathematically grounded dissimilarity metric, improving faithfulness over existing methods.

Findings

CID improves faithfulness metrics for explanations

It offers a complementary perspective to existing explainers

The method is validated against established local importance methods

Abstract

Assessing the importance of individual features in Machine Learning is critical to understand the model's decision-making process. While numerous methods exist, the lack of a definitive ground truth for comparison highlights the need for alternative, well-founded measures. This paper introduces a novel post-hoc local feature importance method called Counterfactual Importance Distribution (CID). We generate two sets of positive and negative counterfactuals, model their distributions using Kernel Density Estimation, and rank features based on a distributional dissimilarity measure. This measure, grounded in a rigorous mathematical framework, satisfies key properties required to function as a valid metric. We showcase the effectiveness of our method by comparing with well-established local feature importance explainers. Our method not only offers complementary perspectives to existing approaches, but also improves performance on faithfulness metrics (both for comprehensiveness and sufficiency), resulting in more faithful explanations of the system. These results highlight its potential as a valuable tool for model analysis.