Research map / Learning theory and algorithms / Kernels and regression theory
Hunting for Discriminatory Proxies in Linear Regression Models
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 |
|---|---|---|---|
| Use the approximate proxy search instead of the exact search | SSL score linear regression approximation · Chicago SSL, race proxy; association thresholds 0.01 to 0.07 | influence estimate: Actual influence: 0.8816; 0.2263; 0.1090; 0.0427*; 0.0065*; 0.0028; 0.0000. Approximate influence: 1.6933; 0.6683; 0.382 | Mixed |
| Exempt the gang-membership indicator | SSL score linear regression approximation · Chicago SSL, race | association: 0.033 | Hurt |
| Combine multiple input variables in a proxy rather than use only the strongest individually associated input | C&C violent-crime linear regression · Communities and Crimes | influence: 0.34 | Helped |
| Use the C&C-trained regression model rather than the dataset response alone | C&C violent-crime prediction · Communities and Crimes | association with protected attribute: 0.48 -> 0.65 | Hurt |
Closest papers by meaning
- Regularizing towards Causal Invariance: Linear Models with Proxies (ICML 2021 · similarity 0.85)
- Reconciling Predictive and Statistical Parity: A Causal Approach (AAAI 2024 · similarity 0.84)
- FairPFN: A Tabular Foundation Model for Causal Fairness (ICML 2025 · similarity 0.83)
- Arbitrariness and Social Prediction: The Confounding Role of Variance in Fair Classification (AAAI 2024 · similarity 0.83)
- Mind the Gap: A Causal Perspective on Bias Amplification in Prediction & Decision-Making (NeurIPS 2024 · similarity 0.82)
- Fairness-Accuracy Trade-Offs: A Causal Perspective (AAAI 2025 · similarity 0.82)
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.