Research map / Federated learning and privacy / Differential privacy
Hypothesis Testing for Differentially Private Linear Regression
Daniel Alabi, Salil Vadhan
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 |
|---|---|---|---|
| Increase sample size | Differentially private tests · Synthetic and Opportunity Insights data | power | Improved |
| Relax the privacy parameter | Differentially private tests · Synthetic and Opportunity Insights data | power | Improved |
| Use the DP F-statistic instead of the DP confidence-interval test | Linear relationship testing · Synthetic and Opportunity Insights data | power | Improved |
| Use the Bernoulli nonparametric tester instead of the DP F-statistic | Linear relationship testing · Synthetic data | power | Mixed |
| Use the Kruskal-Wallis nonparametric tester instead of the DP F-statistic | Mixture testing · Synthetic data | power | Mixed |
| Increase the variance of the independent variable | Mixture testing · Synthetic data | power | Improved |
| Increase the difference between group slopes | Mixture tests · Synthetic data | power | Improved |
| Make group sizes more uneven | Mixture tests · Synthetic data | power | Hurt |
| Increase the slope or decrease dependent-variable noise | DP OLS linear relationship tester · Synthetic data | power | Improved |
| Use DP sufficient statistics and Monte Carlo parametric bootstrap for the F-statistic test | Differentially private hypothesis tests · General linear model | significance | No clear effect |
Closest papers by meaning
- Differentially Private Hypothesis Testing for Linear Regression (JMLR 2023 · similarity 0.98)
- The Test of Tests: A Framework for Differentially Private Hypothesis Testing (ICML 2023 · similarity 0.89)
- Differentially private methods for managing model uncertainty in linear regression (JMLR 2024 · similarity 0.89)
- Priv’IT: Private and Sample Efficient Identity Testing (ICML 2017 · similarity 0.88)
- Locally Private Hypothesis Testing (ICML 2018 · similarity 0.87)
- Minimax Optimal Two-Sample Testing under Local Differential Privacy (NeurIPS 2026 · similarity 0.85)
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.
