Measure Twice, Cut Once: Quantifying Bias and Fairness in Deep Neural Networks

TL;DR

This paper introduces two new metrics, CEV and SDE, for quantifying bias and fairness in multi-class neural network classifiers, addressing a gap in bias evaluation methods.

Contribution

The paper proposes effective, simple metrics for measuring class-wise bias and fairness in multi-class classifiers, filling a notable gap in bias evaluation tools.

Findings

CEV and SDE effectively quantify bias in multi-class models.

Metrics can distinguish bias levels between different models.

Application demonstrates practical utility in fairness assessment.

Abstract

Algorithmic bias is of increasing concern, both to the research community, and society at large. Bias in AI is more abstract and unintuitive than traditional forms of discrimination and can be more difficult to detect and mitigate. A clear gap exists in the current literature on evaluating the relative bias in the performance of multi-class classifiers. In this work, we propose two simple yet effective metrics, Combined Error Variance (CEV) and Symmetric Distance Error (SDE), to quantitatively evaluate the class-wise bias of two models in comparison to one another. By evaluating the performance of these new metrics and by demonstrating their practical application, we show that they can be used to measure fairness as well as bias. These demonstrations show that our metrics can address specific needs for measuring bias in multi-class classification.