A Flexible Fairness Framework with Surrogate Loss Reweighting for Addressing Sociodemographic Disparities

2025/03/21 by Xu, Wen, Dolatabadi, Elham
#FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2503.16836

Abstract

This paper presents a new algorithmic fairness framework called \boldsymbolα-\boldsymbolβ Fair Machine Learning (\boldsymbolα-\boldsymbolβ FML), designed to optimize fairness levels across sociodemographic attributes. Our framework employs a new family of surrogate loss functions, paired with loss reweighting techniques, allowing precise control over fairness-accuracy trade-offs through tunable hyperparameters \boldsymbolα and \boldsymbolβ. To efficiently solve the learning objective, we propose Parallel Stochastic Gradient Descent with Surrogate Loss (P-SGD-S) and establish convergence guarantees for both convex and nonconvex loss functions. Experimental results demonstrate that our framework improves overall accuracy while reducing fairness violations, offering a smooth trade-off between standard empirical risk minimization and strict minimax fairness. Results across multiple datasets confirm its adaptability, ensuring fairness improvements without excessive performance degradation.

Related

• Understanding Fairness Surrogate Functions in Algorithmic Fairness
• SLIDE: a surrogate fairness constraint to ensure fairness consistency
• Metrizing Fairness
• Accuracy and Fairness Trade-offs in Machine Learning: A Stochastic Multi-Objective Approach
• The Sharpe predictor for fairness in machine learning