AI Paper Atlas

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

Simplifying Neural Network Training Under Class Imbalance

Ravid Shwartz-Ziv, Micah Goldblum, Yucen Lily Li, C. Bayan Bruss, Andrew Gordon Wilson

NeurIPS 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 8 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 Joint-SSL, SAM-A and M2m to M2mM2m · CIFAR-10 ratio 0.01accuracy: 84.5 -> 86.0Improved
add Joint-SSL, SAM-A and M2m to ERMERM · CIFAR-100 ratio 0.01accuracy: 47.7 -> 48.9Improved
add joint self-supervised VICReg loss (ablation)ERM · CIFAR-10 ratio 0.01accuracy: 84.9 -> 85.2Improved
use SAM-A alone instead of ERM (ablation)ERM · CIFAR-10 ratio 0.02accuracy: 84.2 -> 83.3Hurt
reweighting instead of ERMERM · CIFAR-10 ratio 0.01accuracy: 84.9 -> 81.8Hurt
resampling instead of ERMERM · CIFAR-10 ratio 0.01accuracy: 84.9 -> 82.1Hurt
add SAM-A on top of Joint-SSL (ablation)Joint-SSL · APTOS 2019 Blindnessaccuracy: 89.7 -> 90.1Improved
add modified label smoothing on top of Joint-SSL + SAM-A (ablation)Joint-SSL + SAM-A · SIIM-ISIC Melanomaaccuracy: 96.0 -> 96.2Improved
add modified label smoothing on top of Joint-SSL + SAM-A (ablation)Joint-SSL + SAM-A · EuroSATaccuracy: 99.1 -> 99.3Improved
add SAM-A and modified label smoothing to MLPMLP · Ottoaccuracy: 83.0 -> 83.2Improved
add SAM-A and modified label smoothing to MLPMLP · CoverTypeaccuracy: 97.5 -> 97.6Improved
add Joint-SSL + SAM-A to ERMERM · CIFAR-10 neural collapse test accuracyaccuracy: 84.58 -> 84.80Improved

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