AI Paper Atlas

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

AutoBalance: Optimized Loss Functions for Imbalanced Data

Mingchen Li, Xuechen Zhang, Christos Thrampoulidis, Jiasi Chen, Samet Oymak

NeurIPS 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Group-robust classification · 28 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
Jointly tune l and Δ, initialized with LA loss, rather than use LA loss alone (ablation)AutoBalance · CIFAR100-LTbalanced error: 58.96 -> 56.70Improved
Jointly tune l and Δ, initialized with LA loss, rather than use LA loss alone (ablation)AutoBalance · ImageNet-LTbalanced error: 52.46 -> 50.91Improved
Jointly tune l and Δ, initialized with LA loss, rather than use LA loss alone (ablation)AutoBalance · iNaturalist-2018balanced error: 34.06 -> 33.25Improved
Jointly tune l and Δ rather than tune l alone (ablation)AutoBalance · CIFAR10-LTbalanced error: 23.02 -> 21.39Improved
Jointly tune l and Δ rather than tune l alone (ablation)AutoBalance · CIFAR100-LTbalanced error: 58.71 -> 56.84Improved
Jointly tune l and Δ rather than tune Δ alone (ablation)AutoBalance · ImageNet-LTbalanced error: 53.02 -> 51.74Improved
Jointly tune l and Δ rather than tune Δ alone (ablation)AutoBalance · iNaturalist-2018balanced error: 34.37 -> 33.41Improved
Initialize joint l and Δ search with LA loss rather than balanced CE (ablation)AutoBalance · ImageNet-LTbalanced error: 51.74 -> 50.91Improved
Use AutoBalance rather than the strongest reported baselineAutoBalance · Modified WaterbirdDEO: 6.91 -> 4.25Improved
Use AutoBalance rather than the strongest reported baselineAutoBalance · Modified Waterbirdbalanced error: 16.47 -> 15.13Improved

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