AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Closing the Approximation Gap of Partial AUC Optimization: A Tale of Two Formulations

Yangbangyan Jiang, Qianqian Xu, Huiyang Shao, Zhiyong Yang, Shilong Bao, Xiaochun Cao, Qingming Huang

TPAMI 2025 · Learning theory and algorithms › Prediction and decision losses · approach: Distributionally robust optimization · 2 citations (OpenAlex)

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 changedWhereMeasuredEffect
use unbiased reformulation (UPAUCI) instead of vanishing-gap approximation (PAUCI)PAUCI · iNaturalist2021 OPAUC FPR<=0.3OPAUC: 0.9348 -> 0.9587Improved
use unbiased reformulation (UPAUCI) instead of PAUCIPAUCI · iNaturalist2021 TPAUC (0.3,0.3)TPAUC: 0.8219 -> 0.8458Improved
use unbiased reformulation (UPAUCI) instead of PAUCIPAUCI · Tiny-ImageNet-200-LT-1 OPAUC FPR<=0.3OPAUC: 0.8267 -> 0.8559Improved
use unbiased reformulation (UPAUCI) instead of PAUCIPAUCI · Tiny-ImageNet-200-LT-1 TPAUC (0.3,0.3)TPAUC: 0.4606 -> 0.4927Improved
use unbiased reformulation (UPAUCI) instead of PAUCIPAUCI · CIFAR-10-LT-3 OPAUC FPR<=0.3OPAUC: 0.7746 -> 0.8061Improved
use unbiased reformulation (UPAUCI) instead of PAUCIPAUCI · CIFAR-100-LT-3 TPAUC (0.3,0.3)TPAUC: 0.5027 -> 0.5465Improved
use unbiased reformulation (UPAUCI) instead of PAUCIPAUCI · CIFAR-10-LT-2 TPAUC (0.3,0.3)TPAUC: 0.9021 -> 0.9015No clear effect
use instance-wise PAUCI instead of AUC optimization (AUC-M)AUC-M · Tiny-ImageNet-200-LT-1 OPAUC FPR<=0.3OPAUC: 0.8102 -> 0.8267Improved
use unbiased UPAUCI instead of biased approximations AUC-polyAUC-poly · Tiny-ImageNet-200-LT-1 OPAUC FPR<=0.3OPAUC: 0.8185 -> 0.8559Improved
use UPAUCI instead of SOPA on CIFAR-100-LT-1 OPAUC FPR<=0.5SOPA · CIFAR-100-LT-1 OPAUC FPR<=0.5OPAUC: 0.9275 -> 0.9221Hurt
initialize c larger (1 instead of 0.2) (ablation)UPAUCI · CIFAR-10-LT-1 OPAUC FPR<=0.5OPAUC: gap approaching 0.1Improved
add small strongly-concave squared term with omega (vs omega=0) (ablation)UPAUCI · CIFAR-10-LT-1OPAUC/TPAUCImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy