AI Paper Atlas

Research map / Learning theory and algorithms / Fairness

A Game-Theoretic Framework for Measuring and Explaining Metric Compatibility in Fair Machine Learning

Lingfeng Zhang, Jingran Yang, Zhaohui Wang, Min Zhang, Zhang Qing

ICML 2026 · Learning theory and algorithms › Fairness · approach: Fairness auditing and regularization

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
apply Fair-SMOTE debiasingVanilla models · Compatibility under debiasingOverall mean compatibility: 0.2518 -> 0.1405Hurt
apply MAAT debiasingVanilla models · Compatibility under debiasingOverall mean compatibility: 0.2518 -> 0.2219Hurt
apply Fair-SMOTE debiasingVanilla models · Compatibility under debiasingGF-U compatibility: -0.0380 -> -0.0507Hurt
apply Fair-SMOTE debiasingVanilla models · Compatibility under debiasingGF-IF compatibility: 0.41 -> 0.1271Hurt
apply Fair-SMOTE debiasingVanilla models · Compatibility under debiasingU-U compatibility: 0.7454 -> 0.5683Hurt
apply MAAT debiasingVanilla models · Compatibility under debiasingU-U compatibility: 0.7454 -> 0.7823Improved
apply Blindness debiasingVanilla models · Compatibility under debiasingIF-IF compatibility: 0.4154 -> 0.0000Hurt
apply MAAT debiasingVanilla models · Compatibility under debiasingIF-U compatibility: -0.0340 -> -0.0352No clear effect
apply Fair-SMOTE debiasingVanilla models · Compatibility under debiasingIF-U compatibility: -0.0340 -> -0.0241Improved
increase HIFI regularization strength eta beyond about 1000HIFI, LR on Census Income · Census IncomeU-U compatibility: from ~1.0 to ~0.6Hurt
use HIFI with eta below 1000HIFI, LR on Census Income · Census IncomeAOD: AOD reduces by ~40%Improved

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