Research map / Learning theory and algorithms / Fairness
Empirical Risk Minimization Under Fairness Constraints
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 |
|---|---|---|---|
| add fairness constraint (ERM under fairness constraints) to nonlinear SVM, s in x | SVM · Adult | DEO: 0.14 -> 0.01 | Helped |
| add fairness constraint to nonlinear SVM, s in x | SVM · Drug | DEO: 0.38±0.03 -> 0.07±0.05 | Helped |
| add fairness constraint to nonlinear SVM, s in x | SVM · Drug | accuracy: 0.82±0.02 -> 0.80±0.03 | Mixed |
| add fairness constraint to linear SVM, s in x | Linear SVM · Arrhythmia | accuracy: 0.79±0.06 -> 0.79±0.07 | No clear effect |
| ours vs Hardt post-processing, nonlinear | Hardt · Drug | DEO: 0.14±0.08 -> 0.07±0.05 | Helped |
| ours vs Zafar, linear, s not in x | Zafar · Drug | accuracy: 0.66±0.03 -> 0.79±0.02 | Helped |
| apply fairness preprocessing to Lasso | Lasso · Arrhythmia | DEO: 0.07±0.04 -> 0.03±0.02 | Helped |
| apply fairness preprocessing to Lasso on 50% reduced training set | Lasso · Arrhythmia | DEO: 0.09±0.05 -> 0.03±0.04 | Helped |
| include sensitive variable s in x for standard SVM | Naive SVM · Drug | DEO: 0.22±0.04 -> 0.45±0.04 | Hurt |
| set epsilon 0.1 instead of 0 (ablation) | ours nonlinear · all datasets | DEO / accuracy | Mixed |
| reduce training set by 50% | Lasso / SVM · Arrhythmia | DEO: 0.07±0.04 -> 0.09±0.05 for Lasso | Hurt |
Closest papers by meaning
- f-FERM: A Scalable Framework for Robust Fair Empirical Risk Minimization (ICLR 2024 · similarity 0.88)
- Fairness Overfitting in Machine Learning: An Information-Theoretic Perspective (ICML 2025 · similarity 0.87)
- Fairness-aware Bayes Optimal Functional Classification (NeurIPS 2025 · similarity 0.86)
- Kernel-based Equalized Odds: A Quantification of Accuracy-Fairness Trade-off in Fair Representation Learning (NeurIPS 2025 · similarity 0.86)
- Optimal Fair Learning Robust to Adversarial Distribution Shift (ICML 2025 · similarity 0.86)
- Learning with Statistical Equality Constraints (NeurIPS 2025 · similarity 0.86)
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.