🤖 AI Summary
Current model evaluations often rely on aggregate metrics that obscure performance disparities and unfairness across continuous or fine-grained subpopulations. This work proposes FairTree, an algorithm that introduces bias-variance decomposition into fairness auditing for the first time, drawing inspiration from measurement invariance in psychometrics to handle continuous, categorical, and ordinal attributes without requiring discretization. By integrating permutation tests with fluctuation tests, FairTree flexibly models subpopulation performance variation and enables rigorous statistical inference. Empirical results demonstrate that FairTree effectively controls false positive rates, with the fluctuation test exhibiting superior statistical power, and its practical utility is validated on the UCI Adult Census dataset.
📝 Abstract
The evaluation of machine learning models typically relies mainly on performance metrics based on loss functions, which risk to overlook changes in performance in relevant subgroups. Auditing tools such as SliceFinder and SliceLine were proposed to detect such groups, but usually have conceptual disadvantages, such as the inability to directly address continuous covariates. In this paper, we introduce FairTree, a novel algorithm adapted from psychometric invariance testing. Unlike SliceFinder and related algorithms, FairTree directly handles continuous, categorical, and ordinal features without discretization. It further decomposes performance disparities into systematic bias and variance, allowing a categorization of changes in algorithm performance. We propose and evaluate two variations of the algorithm: a permutation-based approach, which is conceptually closer to SliceFinder, and a fluctuation test. Through simulation studies that include a direct comparison with SliceLine, we demonstrate that both approaches have a satisfactory rate of false-positive results, but that the fluctuation approach has relatively higher power. We further illustrate the method on the UCI Adult Census dataset. The proposed algorithms provide a flexible framework for the statistical evaluation of the performance and aspects of fairness of machine learning models in a wide range of applications even in relatively small data.