Research map / Learning theory and algorithms / Prediction and decision losses
Asymptotically Unbiased Instance-wise Regularized Partial AUC Optimization: Theory and Algorithm
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 PAUCI instead of the strongest listed baseline | OPAUC optimization · CIFAR-10-LT Subset 1 | OPAUC: 0.7659 -> 0.7721 | Helped |
| Use PAUCI instead of the strongest listed baseline | OPAUC optimization · CIFAR-100-LT Subset 3 | OPAUC: 0.8575 -> 0.8492 | Hurt |
| Use PAUCI instead of the strongest listed baseline | TPAUC optimization · CIFAR-10-LT Subset 1 | TPAUC: 0.7096 -> 0.7192 | Helped |
| Use PAUCI instead of the strongest listed baseline | TPAUC optimization · CIFAR-100-LT Subset 3 | TPAUC: 0.7583 -> 0.7497 | Hurt |
| Replace pair-wise PAUC formulation with an instance-wise formulation | PAUCI · OPAUC and TPAUC | per-iteration time complexity: O(n_+^B n_-^B) -> O(n_+^B + n_-^B) | Helped |
| Use the proposed stochastic minimax solver | PAUCI · OPAUC and TPAUC | iterations to an ε-stationary point: O(ε^-4) -> O(ε^-3) | Helped |
| Smooth the hinge with softplus | PAUCI objective · OPAUC and TPAUC | approximation bias: O(1/κ) | Hurt |
Closest papers by meaning
- When All We Need is a Piece of the Pie: A Generic Framework for Optimizing Two-way Partial AUC (ICML 2021 · similarity 0.91)
- Semi-Supervised Learning for Maximizing the Partial AUC (AAAI 2020 · similarity 0.90)
- AUC Optimization from Multiple Unlabeled Datasets (AAAI 2024 · similarity 0.86)
- Interplay of ROC and Precision-Recall AUCs: Theoretical Limits and Practical Implications in Binary Classification (ICML 2024 · similarity 0.86)
- Positive-unlabeled AUC Maximization under Covariate Shift (ICML 2025 · similarity 0.85)
- AUC Maximization under Positive Distribution Shift (NeurIPS 2024 · similarity 0.85)
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.