Reducing Class-Wise Performance Disparity via Margin Regularization
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866915764289667072 |
|---|---|
| author | Zhu, Beier Zhao, Kesen Cui, Jiequan Sun, Qianru Zhou, Yuan Yang, Xun Zhang, Hanwang |
| author_facet | Zhu, Beier Zhao, Kesen Cui, Jiequan Sun, Qianru Zhou, Yuan Yang, Xun Zhang, Hanwang |
| contents | Deep neural networks often exhibit substantial disparities in class-wise accuracy, even when trained on class-balanced data, posing concerns for reliable deployment. While prior efforts have explored empirical remedies, a theoretical understanding of such performance disparities in classification remains limited. In this work, we present Margin Regularization for Performance Disparity Reduction (MR$^2$), a theoretically principled regularization for classification by dynamically adjusting margins in both the logit and representation spaces. Our analysis establishes a margin-based, class-sensitive generalization bound that reveals how per-class feature variability contributes to error, motivating the use of larger margins for hard classes. Guided by this insight, MR$^2$ optimizes per-class logit margins proportional to feature spread and penalizes excessive representation margins to enhance intra-class compactness. Experiments on seven datasets, including ImageNet, and diverse pre-trained backbones (MAE, MoCov2, CLIP) demonstrate that MR$^2$ not only improves overall accuracy but also significantly boosts hard class performance without trading off easy classes, thus reducing performance disparity. Code is available at: https://github.com/BeierZhu/MR2 |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_00205 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Reducing Class-Wise Performance Disparity via Margin Regularization Zhu, Beier Zhao, Kesen Cui, Jiequan Sun, Qianru Zhou, Yuan Yang, Xun Zhang, Hanwang Machine Learning Computer Vision and Pattern Recognition Deep neural networks often exhibit substantial disparities in class-wise accuracy, even when trained on class-balanced data, posing concerns for reliable deployment. While prior efforts have explored empirical remedies, a theoretical understanding of such performance disparities in classification remains limited. In this work, we present Margin Regularization for Performance Disparity Reduction (MR$^2$), a theoretically principled regularization for classification by dynamically adjusting margins in both the logit and representation spaces. Our analysis establishes a margin-based, class-sensitive generalization bound that reveals how per-class feature variability contributes to error, motivating the use of larger margins for hard classes. Guided by this insight, MR$^2$ optimizes per-class logit margins proportional to feature spread and penalizes excessive representation margins to enhance intra-class compactness. Experiments on seven datasets, including ImageNet, and diverse pre-trained backbones (MAE, MoCov2, CLIP) demonstrate that MR$^2$ not only improves overall accuracy but also significantly boosts hard class performance without trading off easy classes, thus reducing performance disparity. Code is available at: https://github.com/BeierZhu/MR2 |
| title | Reducing Class-Wise Performance Disparity via Margin Regularization |
| topic | Machine Learning Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.00205 |