[2603.21393] A Generalised Exponentiated Gradient Approach to Enhance Fairness in Binary and Multi-class Classification Tasks

Computer Science > Machine Learning arXiv:2603.21393 (cs) [Submitted on 22 Mar 2026] Title:A Generalised Exponentiated Gradient Approach to Enhance Fairness in Binary and Multi-class Classification Tasks Authors:Maryam Boubekraoui, Giordano d'Aloisio, Antinisca Di Marco View a PDF of the paper titled A Generalised Exponentiated Gradient Approach to Enhance Fairness in Binary and Multi-class Classification Tasks, by Maryam Boubekraoui and 2 other authors View PDF Abstract:The widespread use of AI and ML models in sensitive areas raises significant concerns about fairness. While the research community has introduced various methods for bias mitigation in binary classification tasks, the issue remains under-explored in multi-class classification settings. To address this limitation, in this paper, we first formulate the problem of fair learning in multi-class classification as a multi-objective problem between effectiveness (i.e., prediction correctness) and multiple linear fairness constraints. Next, we propose a Generalised Exponentiated Gradient (GEG) algorithm to solve this task. GEG is an in-processing algorithm that enhances fairness in binary and multi-class classification settings under multiple fairness definitions. We conduct an extensive empirical evaluation of GEG against six baselines across seven multi-class and three binary datasets, using four widely adopted effectiveness metrics and three fairness definitions. GEG overcomes existing baselines, with fairness i...