A Distributionally Robust Approach to Fair Classification

2020/07/18 by Bahar Taşkesen, Taskesen, Bahar, Viet Anh Nguyen +5 · 5 citations
Economics, Econometrics and Finance · #Health Systems, Economic Evaluations, Quality of Life

paper · pdf · doi:10.48550/arxiv.2007.09530

Abstract

We propose a distributionally robust logistic regression model with an unfairness penalty that prevents discrimination with respect to sensitive attributes such as gender or ethnicity. This model is equivalent to a tractable convex optimization problem if a Wasserstein ball centered at the empirical distribution on the training data is used to model distributional uncertainty and if a new convex unfairness measure is used to incentivize equalized opportunities. We demonstrate that the resulting classifier improves fairness at a marginal loss of predictive accuracy on both synthetic and real datasets. We also derive linear programming-based confidence bounds on the level of unfairness of any pre-trained classifier by leveraging techniques from optimal uncertainty quantification over Wasserstein balls.

Citations

• Equality of Opportunity in Supervised Learning
• Fairness in Criminal Justice Risk Assessments: The State of the Art
• On Fairness and Calibration
• Data-driven distributionally robust optimization using the Wasserstein metric: performance guarantees and tractable reformulations

Cited by

• Algorithmic Bias and Data Bias: Understanding the Relation between Distributionally Robust Optimization and Data Curation
• Wasserstein projection distance for fairness testing of regression models
• From Fragile to Certified: Wasserstein Audits of Group Fairness Under Distribution Shift
• Robust Fairness-aware Learning Under Sample Selection Bias
• Sinkhorn Distributionally Robust Optimization

Related

• Wasserstein Robust Classification with Fairness Constraints
• Fairness Uncertainty Quantification: How certain are you that the model is fair?
• Long Term Fairness for Minority Groups via Performative Distributionally Robust Optimization
• Group Fairness with Uncertainty in Sensitive Attributes
• Distributionally Robust Logistic Regression