AI Paper Atlas

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

When False Positive is Intolerant: End-to-End Optimization with Low FPR for Multipartite Ranking

Peisong Wen, Qianqian Xu, Zhiyong Yang, Yuan He, Qingming Huang

NeurIPS 2021 · Learning theory and algorithms › Prediction and decision losses · approach: Ranking-based loss optimization · 0 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
CBA-CR instead of cross-entropy trainingReXNet200 · DR TPR@FPR=0.003TPR (%): 35.2 -> 39.1Improved
CBA-CR instead of cross-entropy trainingReXNet200 · UTKFace TPR@FPR=0.003TPR (%): 33.1 -> 44.9Improved
CBA-CR instead of cross-entropy trainingReXNet200 · AgePrediction TPR@FPR=0.003TPR (%): 11.3 -> 16.0Improved
CBA-CR instead of Eban's methodReXNet200 · NSFW TPR@FPR=0.003TPR (%): 56.3 -> 58.1Improved
larger FPR budget reduces gain over best competitor on DRCBA-CR · DRTPR gain over best competitor: +2.3, +2.4, +0.6 at alpha 0.003, 0.01, 0.03Mixed
existing ranking/ordinal methods versus CE baselineNNRank, F&H, SoftLabel, DDAG · DR/UTKFace/NSFW/AgePredictionTPR@FPRNo clear effect
CBA-CR at lower FPR gives larger improvementsCBA-CR · all datasetsTPR gain: UTKFace 33.1 -> 44.9 at 0.003 vs 80.9 -> 82.9 at 0.03Improved
CBA-CR versus CE on AUCReXNet200 · DR AUCAUC (%): 84.4 -> 86.3Improved
CBA-CR versus SoftLabel on AUCReXNet200 · UTKFace AUCAUC (%): 96.0 -> 95.9No clear effect
add AUC loss to F&HF&H · UTKFace TPR@FPR=0.003TPR (%): 37.5 -> 42.1Improved
cache size Nc too small or too large (ablation)CBA-CR · DR sweep Nc in {64..2048}TPR@FPRHurt
cache size effect at alpha 0.003 (ablation)CBA-CR · DRTPR@FPR=0.003No clear effect

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