Research map / Learning theory and algorithms / Prediction and decision losses
Weighted ROC Curve in Cost Space: Extending AUC to Cost-Sensitive Learning
Huiyang Shao, Qianqian Xu, Zhiyong Yang, Peisong Wen, Gao Peifeng, Qingming Huang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use WAUC-Gau bilevel cost-sensitive training instead of ECL | cost-sensitive classifier · CIFAR-10-LT Subset1 | WAUC: 0.589 -> 0.679 | Improved |
| Use WAUC-Gau instead of NWAUC | cost-sensitive classifier · CIFAR-100-LT Subset3 | WAUC: 0.631 -> 0.728 | Improved |
| Use WAUC-Log instead of ECL | cost-sensitive classifier · CIFAR-100-LT Subset1 | bLCOST: 0.017 -> 0.012 | Improved |
| Use WAUC-Gau instead of AdaCOS | cost-sensitive classifier · Jane Street Market Prediction | Profit: 5839 ± 34.92 -> 6526 ± 15.98 | Improved |
| Use WAUC-Gau instead of BCE | cost-sensitive classifier · Jane Street Market Prediction | Profit: 4955 ± 20.14 -> 6526 ± 15.98 | Improved |
| Use WAUC-Gau instead of ExAUC | cost-sensitive classifier · Jane Street Market Prediction | AUC-hat: 0.539 ± .004 -> 0.526 ± .002 | No clear effect |
| Use WAUC-Log instead of ExAUC (AUC-oriented method) | cost-sensitive classifier · CIFAR-100-LT Subset1 | AUC: 0.933 -> 0.906 | Mixed |
| Use two-stage decision (train classifier or AUC, then pick cost threshold) instead of one-stage WAUC learning | BCE and AUC methods · cost-sensitive benchmarks | WAUC and bLCOST | Hurt |
| Choose smaller T in the inner threshold estimation (ablation) | WAUC-Gau · CIFAR benchmarks | WAUC | Improved |
| Choose larger T (ablation) | WAUC-Gau · CIFAR benchmarks | variance of WAUC | Improved |
| Vary kernel bandwidth in KDE (ablation) | WAUC-Gau · CIFAR benchmarks | WAUC and cost | Mixed |
Closest papers by meaning
- Learning With Multiclass AUC: Theory and Algorithms (TPAMI 2021 · similarity 0.89)
- Cost-Aware AUC Optimization via Adaptive Kernel Density Estimation (TPAMI 2025 · similarity 0.89)
- AUC Optimization from Multiple Unlabeled Datasets (AAAI 2024 · similarity 0.85)
- Positive-unlabeled AUC Maximization under Covariate Shift (ICML 2025 · similarity 0.85)
- Interplay of ROC and Precision-Recall AUCs: Theoretical Limits and Practical Implications in Binary Classification (ICML 2024 · similarity 0.85)
- The VOROS: Lifting ROC Curves to 3D to Summarize Unbalanced Classifier Performance (AAAI 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
