AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

A Unified Generalization Analysis of Re-Weighting and Logit-Adjustment for Imbalanced Learning

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

NeurIPS 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Group-robust classification · 1 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
add DRW re-weighting to CE (ablation)CE · CIFAR-10 LT rho=100balanced accuracy: 71.5 -> 75.8Improved
replace DRW with adaptive ADRW re-weighting (ablation)CE+DRW · CIFAR-10 LT rho=100balanced accuracy: 75.8 -> 78.6Improved
replace DRW with ADRW (ablation)LDAM+DRW · CIFAR-10 LT rho=100balanced accuracy: 77.7 -> 79.1Improved
add DRW to VS loss (multiplicative logit adjustment) (ablation)VS · CIFAR-100 LT rho=100balanced accuracy: 41.8 -> 41.3Hurt
add DRW to VS loss (ablation)VS · CIFAR-100 Step rho=10balanced accuracy: 59.9 -> 58.0Hurt
add DRW to VS loss (ablation)VS · CIFAR-10 LT rho=100balanced accuracy: 78.8 -> 80.1Improved
add TLA (additive logit adjustment) to VS+DRW (ablation)VS+DRW · CIFAR-100 LT rho=100balanced accuracy: 41.3 -> 43.0Improved
replace DRW with ADRW in VS+TLA (ablation)VS+TLA · CIFAR-10 Step rho=100balanced accuracy: 80.0 -> 80.9Improved
add SAM to VS+TLA+ADRW (ablation)VS+TLA+ADRW · CIFAR-10 LT rho=100balanced accuracy: 81.1 -> 83.6Improved
add SAM to VS+TLA+ADRW (ablation)VS+TLA+ADRW · CIFAR-100 LT rho=100balanced accuracy: 43.4 -> 46.4Improved
use CE+ADRW instead of LDAM+ADRW at rho=10 (ablation)CE+ADRW · CIFAR-100 LT rho=10balanced accuracy: 58.0 -> 58.3Mixed
use class-balanced (CB) loss instead of DRW (ablation)CE · ImageNet-LTbalanced accuracy (All): 44.4 -> 33.2Hurt

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