atlas

Research map / Learning theory and algorithms / Fairness

On Learning Fairness and Accuracy on Multiple Subgroups

NeurIPS 2022 · Learning theory and algorithms › Fairness · approach: Fairness-constrained classification · 16 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
use bilevel FAMS instead of ERMDistilBERT MLP · Amazon reviewgroup sufficiency gap: 3.0% improvementHelped
use bilevel FAMS instead of ERMDistilBERT MLP · Amazon reviewaccuracyHurt
use FAMS instead of baselinesDistilBERT MLP · Toxic comments (race)group sufficiency gap: 3.0-3.5% improvementHelped
use FAMS instead of ERM2-layer MLP · Adultgroup sufficiency gap: 1.096 -> 0.684Helped
use FAMS instead of ERM2-layer MLP · Adultaccuracy: 83.38 -> 83.33No clear effect
use FAMS instead of ERMResNet-18 · CelebAgroup sufficiency gap: 7.13 -> 5.29Helped
use FAMS instead of ERMResNet-18 · CelebAaccuracy: 79.67 -> 79.87No clear effect
set lambda near 0 (no prior regularization in lower level) (ablation)FAMS · Amazon reviewaccuracy and group sufficiency gapHurt
set very large lambda (50) (ablation)FAMS · Amazon reviewaccuracy: 55%Hurt
lambda in [0.2,1] (ablation)FAMS · Amazon reviewaccuracy and group sufficiency gapHelped

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