Research map / Learning theory and algorithms / Fairness
Bounding and Approximating Intersectional Fairness through Marginal Fairness
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 |
|---|---|---|---|
| Group protected attributes using the greedy feasible-partition heuristic | Marginal approximation UFIind · US Census 1990 subsets (D_i) | relative expected L2 error | Helped |
| Use the ungrouped marginal approximation instead of the grouped partition approximation (ablation) | UFIind · US Census 1990 subsets (D_i) | relative expected L2 error | Hurt |
| Use the Bayesian estimator UFIB rather than the marginal approximations | Intersectional unfairness estimation · US Census 1990 subsets (D_i) | relative expected L2 error | Hurt |
| Use the Bayesian estimator UFIB rather than the smax estimator | Independence-departure estimation · Synthetic distributions P_i | relative expected L2 error | Hurt |
| Vary the Bayesian prior parameter alpha (ablation) | Bayesian estimator UFIB · US Census 1990 subsets (D_i) and synthetic distributions P_i | relative expected L2 error | Mixed |
| Increase the grouping count threshold tau (ablation) | Partitioned marginal approximation UFIind[(q*)] · US Census 1990 subsets (D_i) | approximation error | Mixed |
| Use the theoretical high-probability bounds | Intersectional fairness estimation · US Census 1990 subsets (D_i) | bound tightness | Hurt |
| Use ungrouped marginal approximation on the synthetic distributions (ablation) | UFIind · Synthetic distributions P_i | approximation to intersectional unfairness | Hurt |
Closest papers by meaning
- Intersectional Fairness via Mixed-Integer Optimization (NeurIPS 2026 · similarity 0.90)
- Size-adaptive Hypothesis Testing for Fairness (NeurIPS 2025 · similarity 0.88)
- On the Maximal Local Disparity of Fairness-Aware Classifiers (ICML 2024 · similarity 0.87)
- Intersectional Unfairness Discovery (ICML 2024 · similarity 0.87)
- Fairness Overfitting in Machine Learning: An Information-Theoretic Perspective (ICML 2025 · similarity 0.87)
- Searching for Unfairness in Algorithms’ Outputs: Novel Tests and Insights (AAAI 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.