Robust Long-Tailed Learning via Label-Aware Bounded CVaR
This work addresses imbalanced data issues in classification, which is a common problem in real-world applications, but it appears incremental as it builds on existing CVaR methods with modifications.
The paper tackles the problem of long-tailed learning in classification, where models perform poorly on minority classes, by proposing two novel loss functions based on Conditional Value at Risk (CVaR) with theoretical guarantees, achieving improved performance on real-world datasets.
Data in the real-world classification problems are always imbalanced or long-tailed, wherein the majority classes have the most of the samples that dominate the model training. In such setting, the naive model tends to have poor performance on the minority classes. Previously, a variety of loss modifications have been proposed to address the long-tailed leaning problem, while these methods either treat the samples in the same class indiscriminatingly or lack a theoretical guarantee. In this paper, we propose two novel approaches based on CVaR (Conditional Value at Risk) to improve the performance of long-tailed learning with a solid theoretical ground. Specifically, we firstly introduce a Label-Aware Bounded CVaR (LAB-CVaR) loss to overcome the pessimistic result of the original CVaR, and further design the optimal weight bounds for LAB-CVaR theoretically. Based on LAB-CVaR, we additionally propose a LAB-CVaR with logit adjustment (LAB-CVaR-logit) loss to stabilize the optimization process, where we also offer the theoretical support. Extensive experiments on real-world datasets with long-tailed label distributions verify the superiority of our proposed methods.