AI Paper Atlas

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

Optimizing Partial Area Under the Top-K Curve: Theory and Practice

Zitai Wang, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao, Qingming Huang

TPAMI 2022 · Learning theory and algorithms › Prediction and decision losses · approach: Ranking-based loss optimization

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
optimize AUTKC (Exp surrogate) instead of cross-entropyResNet-18 · Cifar-100 (K=3)AUTKC: 71.81 -> 73.60Improved
optimize AUTKC (Exp surrogate) instead of cross-entropyResNet-18 · Tiny-imagenet-200 (K=3)AUTKC: 81.56 -> 82.87Improved
optimize AUTKC (Exp surrogate) instead of cross-entropyResNet-50 features + FC · Places-365 (K=3)AUTKC: 67.72 -> 67.95Improved
optimize AUTKC (Exp surrogate) instead of cross-entropyResNet-18 · Cifar-10 (K=3)AUTKC: 92.87 -> 93.01Improved
use top-k losses instead of cross-entropyResNet-18 · Cifar-10 (K=3)AUTKC: 92.87 -> 92.56Hurt
use top-k loss L3 instead of cross-entropyResNet-18 · Cifar-100 (K=3)AUTKC: 71.81 -> 71.98Improved
use hinge surrogate instead of exp surrogate in AUTKCAUTKC · Cifar-100 (K=3)AUTKC: 73.60 -> 73.29Hurt
use exp surrogate instead of hinge surrogate in AUTKCAUTKC · Tiny-imagenet-200 (K=3)AUTKC: 82.20 -> 82.87Improved
use AUTKC-Hinge instead of traditional hinge lossResNet-18 · Cifar-100 (K=3)AUTKC: 71.33 -> 73.29Improved
use AUTKC-Logit instead of CEResNet-18 · Cifar-10 (K=3)AUTKC: 92.87 -> 93.05Improved

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