FlipTest: Fairness Testing via Optimal Transport

Abstract

We present FlipTest, a black-box technique for uncovering discrimination in classifiers. FlipTest is motivated by the intuitive question: had an individual been of a different protected status, would the model have treated them differently? Rather than relying on causal information to answer this question, FlipTest leverages optimal transport to match individuals in different protected groups, creating similar pairs of in-distribution samples. We show how to use these instances to detect discrimination by constructing a "flipset": the set of individuals whose classifier output changes post-translation, which corresponds to the set of people who may be harmed because of their group membership. To shed light on why the model treats a given subgroup differently, FlipTest produces a "transparency report": a ranking of features that are most associated with the model's behavior on the flipset. Evaluating the approach on three case studies, we show that this provides a computationally inexpensive way to identify subgroups that may be harmed by model discrimination, including in cases where the model satisfies group fairness criteria.

Accepted to ACM FAT* 2020; The first two authors contributed equally

References in corpus (8)

• Actionable Recourse in Linear Classification
• Fairness Testing: Testing Software for Discrimination
• Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness
• Fairness Constraints: Mechanisms for Fair Classification
• Near-linear time approximation algorithms for optimal transport via Sinkhorn iteration
• AI Fairness 360: An Extensible Toolkit for Detecting, Understanding, and Mitigating Unwanted Algorithmic Bias
• Fairness in Learning: Classic and Contextual Bandits
• Hunting for Discriminatory Proxies in Linear Regression Models

Cited by in corpus (12)

• A Clarification of the Nuances in the Fairness Metrics Landscape
• A Critical Survey on Fairness Benefits of Explainable AI
• Amazon SageMaker Clarify: Machine Learning Bias Detection and Explainability in the Cloud
• Latent Imitator: Generating Natural Individual Discriminatory Instances for Black-Box Fairness Testing
• Specification Overfitting in Artificial Intelligence
• FairCanary: Rapid Continuous Explainable Fairness
• FairBalance: How to Achieve Equalized Odds With Data Pre-processing
• Counterfactual Situation Testing: Uncovering Discrimination under Fairness given the Difference
• Locating disparities in machine learning
• Algorithmic Fairness Datasets: the Story so Far
• Constrained Approximate Optimal Transport Maps
• Mutatis Mutandis: Revisiting the Comparator in Discrimination Testing