Robust Learning from Noisily Labeled Long-Tailed Data via Fairness Regularizer
Wei, Jiaheng Zhu, Zhaowei Niu, Gang Liu, Tongliang Liu, Sijia Sugiyama, Masashi Liu, Yang
Wei, Jiaheng
Zhu, Zhaowei
Niu, Gang
Liu, Tongliang
Liu, Sijia
Sugiyama, Masashi
Liu, Yang
Supervisor
Department
Machine Learning
Embargo End Date
Type
Conference proceeding
Date
License
Language
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
Both long-tailed and noisily labeled data frequently appear in real-world applications and impose significant challenges for learning. Most prior works treat either problem in an isolated way and do not explicitly consider the coupling effects of the two. Our empirical observation reveals that such solutions fail to consistently improve the learning when the dataset is long-tailed with label noise. Moreover, with the presence of label noise, existing methods do not observe universal improvements across different sub-populations; in other words, some sub-populations enjoyed the benefits of improved accuracy at the cost of hurting others. Based on these observations, we introduce the Fairness Regularizer (FR), inspired by regularizing the performance gap between any two sub-populations. We show that the introduced fairness regularizer improves the performances of sub-populations on the tail and the overall learning performance. Extensive experiments demonstrate the effectiveness of the proposed solution when complemented with certain existing popular robust or class-balanced methods.
Citation
J. Wei, Z. Zhu, G. Niu, T. Liu, S. Liu, M. Sugiyama , et al., "Robust Learning from Noisily Labeled Long-Tailed Data via Fairness Regularizer," 2026, pp. 35847-35856.
Source
Proceedings of the AAAI Conference on Artificial Intelligence
Conference
Fortieth AAAI Conference on Artificial Intelligence
Keywords
46 Information and Computing Sciences, 4611 Machine Learning
Subjects
Source
Fortieth AAAI Conference on Artificial Intelligence
Publisher
Association for the Advancement of Artificial Intelligence