Research map / Learning theory and algorithms / Fairness
Interpretable algorithmic fairness in structured and unstructured data
Hari Bandi, Dimitris Bertsimas, Thodoris Koukouvinos, Sofie Kupiec
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add meritocracy constraint with delta = 0.1 (ablation) | label-flipping method · LSAC | Merit(lsat): 0.181 -> 0.089 | Improved |
| add meritocracy constraint with delta = 0.1 (ablation) | label-flipping method · LSAC | SPD: 0.011 -> 0.072 | Hurt |
| add meritocracy constraint with delta = 0.1 (ablation) | label-flipping method · Crime | Merit(pct-under-pov): 0.037 -> 0.023 | Improved |
| add meritocracy constraint with delta = 0.1 (ablation) | label-flipping method · COMPAS | Merit(priors): 0.076 -> 0.034 | Improved |
| add meritocracy constraint with delta = 0.1 (ablation) | label-flipping method · German credit | Merit(credit): 0.019 -> 0.015 | Improved |
| relax meritocracy tolerance from delta = 0.1 to delta = 0.5 (ablation) | label-flipping method · LSAC | SPD: 0.072 -> 0.026 | Improved |
| apply label-flipping method (no merit) over nominal model | ResNet-18 / tabular classifiers · LSAC | SPD: 0.299 -> 0.011 | Improved |
| apply label-flipping method (no merit) over nominal model | tabular classifier · Crime | accuracy: 0.779 -> 0.678 | Hurt |
| apply label-flipping method over nominal model | ResNet-18 · CelebA Brown Hair | SPD: 0.035 -> 0.011 | Improved |
| apply label-flipping method over nominal model | ResNet-18 · CelebA Blond Hair | SPD: 0.059 -> 0.041 | Improved |
| alleviate bias jointly for race and gender instead of separately (ablation) | label-flipping method · LSAC | SPD(WM,BM): 0.198 -> 0.007 | Improved |
| alleviate bias jointly for race and gender instead of separately (ablation) | label-flipping method · LFW Brown Hair | SPD(WM,BM): 0.174 -> 0.127 | Improved |
Closest papers by meaning
- The Fragility of Fairness: Causal Sensitivity Analysis for Fair Machine Learning (NeurIPS 2024 · similarity 0.86)
- Post-hoc bias scoring is optimal for fair classification (ICLR 2024 · similarity 0.86)
- Classes Are Not Equal: An Empirical Study on Image Recognition Fairness (CVPR 2024 · similarity 0.86)
- Arbitrariness and Social Prediction: The Confounding Role of Variance in Fair Classification (AAAI 2024 · similarity 0.86)
- Utility-Fairness Trade-Offs and How to Find Them (CVPR 2024 · similarity 0.86)
- A Sequentially Fair Mechanism for Multiple Sensitive Attributes (AAAI 2024 · 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.
