Research map / Federated learning and privacy / Differential privacy
Better Private Linear Regression Through Better Private Feature Selection
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 DPKendall feature selection before Tukey regression instead of using no feature selection | Tukey · 25 linear regression datasets, k=5 | datasets with positive R²: 16% -> 56% | Helped |
| Use DPKendall rather than SubLasso for feature selection before Tukey regression | Tukey · 25 linear regression datasets, k=5 | datasets with positive median test R²: 40% -> 56% | Helped |
| Use DPKendall rather than SubLasso for feature selection before BAS regression | BAS · 25 linear regression datasets, k=5 | datasets with positive median test R²: 20% -> 28% | Helped |
| Use DPKendall rather than SubLasso for feature selection before Tukey regression | Tukey · 25 linear regression datasets, k=5 | datasets ranked first: 40% -> 48% | Helped |
| Increase the selected feature count from k=5 to k=10 | Private feature-selecting regression methods · 21 linear regression datasets | total positive R² results across methods: 41 -> 19 | Hurt |
Closest papers by meaning
- On the Computational Complexity of Private High-dimensional Model Selection (NeurIPS 2024 · similarity 0.90)
- Differentially Private High-dimensional Variable Selection via Integer Programming (NeurIPS 2025 · similarity 0.89)
- The importance of feature preprocessing for differentially private linear optimization (ICLR 2024 · similarity 0.89)
- Instance-Specific Asymmetric Sensitivity in Differential Privacy (NeurIPS 2024 · similarity 0.88)
- Private Gradient Descent for Linear Regression: Tighter Error Bounds and Instance-Specific Uncertainty Estimation (ICML 2024 · similarity 0.88)
- Certification for Differentially Private Prediction in Gradient-Based Training (ICML 2025 · similarity 0.88)
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.