Research map / Learning theory and algorithms / Fairness
Minimax AUC Fairness: Efficient Algorithm with Provable Convergence
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Optimize the minimum across all intra- and inter-group AUCs instead of unconstrained overall AUC | AUCMax training on Adult · Adult | Min/Max group AUC: .823±.005 -> .953±.002 | Helped |
| Optimize the minimum across all intra- and inter-group AUCs instead of unconstrained overall AUC | AUCMax training on Bank · Bank | Min/Max group AUC: .780±.018 -> .858±.014 | Helped |
| Optimize the minimum across all intra- and inter-group AUCs instead of unconstrained overall AUC | AUCMax training on Compas · Compas | Overall AUC: .732±.004 -> .741±.004 | Helped |
| Optimize the minimum across all intra- and inter-group AUCs instead of unconstrained overall AUC | AUCMax training on Default · Default | Overall AUC: .763±.005 -> .767±.002 | Helped |
| Optimize the minimum across all intra- and inter-group AUCs instead of unconstrained overall AUC | AUCMax training on Adult · Adult | Overall AUC: .902±.002 -> .901±.004 | Hurt |
| Optimize the minimum across all intra- and inter-group AUCs instead of unconstrained overall AUC | AUCMax training on Bank · Bank | Overall AUC: .910±.002 -> .907±.004 | Hurt |
| Use the minimax AUC objective rather than EqualAUC constraints | Adult · Adult | Overall AUC: .886±.003 -> .901±.004 | Helped |
| Use the minimax AUC objective rather than EqualAUC constraints | Bank · Bank | Overall AUC: .866±.005 -> .907±.004 | Helped |
| Use the minimax AUC objective rather than EqualAUC constraints | Bank · Bank | Min/Max group AUC: .891±.025 -> .858±.014 | Hurt |
| Use the minimax AUC objective rather than EqualAUC constraints | Compas · Compas | Overall AUC: .731±.003 -> .741±.004 | Helped |
| Use the minimax AUC objective rather than EqualAUC constraints | Compas · Compas | Min/Max group AUC: .956±.012 -> .961±.012 | Helped |
| Use only intra-group AUCs, dropping inter-group AUCs (ablation) | MinimaxFairAUC on Adult · Adult | Intra-group AUC fairness | Mixed |
Closest papers by meaning
- Preserving AUC Fairness in Learning with Noisy Protected Groups (ICML 2025 · similarity 0.91)
- FFB: A Fair Fairness Benchmark for In-Processing Group Fairness Methods (ICLR 2024 · similarity 0.87)
- Individual Fairness In Strategic Classification (NeurIPS 2025 · similarity 0.86)
- Enhancing Group Fairness in Online Settings Using Oblique Decision Forests (ICLR 2024 · similarity 0.85)
- Intersectional Fairness via Mixed-Integer Optimization (NeurIPS 2026 · similarity 0.85)
- On the Maximal Local Disparity of Fairness-Aware Classifiers (ICML 2024 · similarity 0.85)
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.