IFFair: Influence Function-driven Sample Reweighting for Fair Classification
Jingran Yang, Min Zhang, Lingfeng Zhang, Zhaohui Wang, Yonggang Zhang
TL;DR
The paper tackles bias in classification by introducing IFFair, a pre-processing method that reweights training samples using influence-function-based group disparity rather than modifying the model or data features. It formalizes group-oriented influence and provides two reweighting variants, Uniform and Diverse, with trade-off constraints to balance fairness and utility. Extensive experiments across LR and DNN models on seven real-world datasets show that IFFair achieves substantial improvements in multiple fairness metrics with minimal or no deterioration in utility, often outperforming existing pre-processing baselines. The approach is shown to be generalizable, scalable across architectures, and capable of achieving favorable fairness-utility trade-offs, underscoring its practical potential for fair classification.
Abstract
Because machine learning has significantly improved efficiency and convenience in the society, it's increasingly used to assist or replace human decision-making. However, the data-based pattern makes related algorithms learn and even exacerbate potential bias in samples, resulting in discriminatory decisions against certain unprivileged groups, depriving them of the rights to equal treatment, thus damaging the social well-being and hindering the development of related applications. Therefore, we propose a pre-processing method IFFair based on the influence function. Compared with other fairness optimization approaches, IFFair only uses the influence disparity of training samples on different groups as a guidance to dynamically adjust the sample weights during training without modifying the network structure, data features and decision boundaries. To evaluate the validity of IFFair, we conduct experiments on multiple real-world datasets and metrics. The experimental results show that our approach mitigates bias of multiple accepted metrics in the classification setting, including demographic parity, equalized odds, equality of opportunity and error rate parity without conflicts. It also demonstrates that IFFair achieves better trade-off between multiple utility and fairness metrics compared with previous pre-processing methods.