arXiv：cs.LG· Ji\v{r}\'i N\v{e}me\v{c}ek, Mark Kozdoba, Illia Kryvoviaz, Tom\'a\v{s} Pevn\'y, Jakub Mare\v{c}ek·· 7 小时前AI 评分32
基于混合整数优化的交叉群体公平性
Intersectional Fairness via Mixed-Integer Optimization
AI 导读
研究者提出一个利用混合整数优化（MIO）训练交叉群体公平且内在可解释分类器的统一框架，并证明 MSD 与 SPSF 两种交叉公平性度量在检测最不公平子群时等价。实验显示该 MIO 算法在发现偏见方面表现更优，可训练出将交叉偏见控制在可接受阈值以下的高性能可解释分类器。该工作为金融、医疗等受监管行业提供了稳健方案，已被 NeurIPS 2026 接收。
正文
Abstract:The deployment of Artificial Intelligence in high-risk domains, such as finance and healthcare, necessitates models that are both fair and transparent. While regulatory frameworks, including the EU's AI Act, mandate bias mitigation, they are deliberately vague about the definition of bias. In line with existing research, we argue that true fairness requires addressing bias at the intersections of protected groups. We propose a unified framework that leverages Mixed-Integer Optimization (MIO) to train intersectionally fair and intrinsically interpretable classifiers. We prove the equivalence of two measures of intersectional fairness (MSD and SPSF) in detecting the most unfair subgroup and empirically demonstrate that our MIO-based algorithm improves performance in finding bias. We train high-performing, interpretable classifiers that bound intersectional bias below an acceptable threshold, offering a robust solution for regulated industries and beyond.
| Comments: | 17 pages, 10 figures, 1 table |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optimization and Control (math.OC); Machine Learning (stat.ML) |
| Cite as: | arXiv:2601.19595 [cs.LG] |
| (or arXiv:2601.19595v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2601.19595 arXiv-issued DOI via DataCite |
|
| Journal reference: | NeurIPS 2026 |
Submission history
From: Jiří Němeček [view email]
[v1]
Tue, 27 Jan 2026 13:29:25 UTC (596 KB)
[v2]
Mon, 5 Oct 2026 21:58:17 UTC (607 KB)
来源：arXiv：cs.LG · arxiv.org