MIFair: A mutual-information framework for intersectionality and multiclass fairness

 Fairness in machine learning remains challenging due to its ethical complexity, the lack of a universal definition, and the need for context-specific bias metrics. Existing methods also remain limited in handling intersectionality, multiclass settings, and broader flexibility and generality. To address these limitations, we introduce MIFair, a unified framework for bias assessment and mitigation based on mutual information. MIFair provides a flexible metric template and an in-processing mitigation method inspired by the Prejudice Remover, defining group fairness as statistical independence between predictionderived variables and sensitive attributes, while establishing equivalences with widely used notions such as independence and separation. MIFair naturally supports intersectionality, complex subgroup structures, and multiclass classification and employs regularization-based training to reduce bias according to the selected metric. Its key advantage is its versatility: it consolidates diverse fairness requirements into a single coherent framework, enabling consistent benchmarking and facilitating practical adoption. Experiments on real-world tabular and image datasets show that MIFair effectively reduces bias, including in previously unaddressed multi-attribute scenarios, while maintaining strong predictive performance across all evaluated settings.