Fair Feature Importance Scores via Feature Occlusion and Permutation

TL;DR

This paper introduces two simple, scalable, and model-agnostic methods to quantify how individual features influence fairness in machine learning models, enhancing interpretability and accountability.

Contribution

The paper presents novel, intervention- and occlusion-based approaches for measuring feature importance specifically for fairness, addressing a gap in existing feature importance metrics.

Findings

Methods are simple and effective across multiple tasks.

Proposed metrics are computationally efficient via minipatch learning.

Tools provide interpretable insights into feature fairness contributions.

Abstract

As machine learning models increasingly impact society, their opaque nature poses challenges to trust and accountability, particularly in fairness contexts. Understanding how individual features influence model outcomes is crucial for building interpretable and equitable models. While feature importance metrics for accuracy are well-established, methods for assessing feature contributions to fairness remain underexplored. We propose two model-agnostic approaches to measure fair feature importance. First, we propose to compare model fairness before and after permuting feature values. This simple intervention-based approach decouples a feature and model predictions to measure its contribution to training. Second, we evaluate the fairness of models trained with and without a given feature. This occlusion-based score enjoys dramatic computational simplification via minipatch learning. Our empirical results reflect the simplicity and effectiveness of our proposed metrics for multiple predictive tasks. Both methods offer simple, scalable, and interpretable solutions to quantify the influence of features on fairness, providing new tools for responsible machine learning development.