AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

The Test of Tests: A Framework for Differentially Private Hypothesis Testing

Zeki Kazan, Kaiyan Shi, Adam Groce, Andrew Bray

ICML 2023 · Federated learning and privacy › Differential privacy · approach: Local-DP and private estimation · 3 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use Awan-Slavkovic UMP binomial test as aggregation step in subsample-and-aggregate (ToT) instead of PB's randomized-response aggregationprivate hypothesis testing framework · Power comparison with PB framework (t-test)power / sample size for 80% power: n=200 vs n=280 for 80% power at eps=1Improved
Use Tulap-based binomial aggregation with tuned number of subsamples m and subtest threshold alpha_0 (optimization favoring large m)Test of Tests · Multivariate normal mean test, d=100, eps=1sample size for 80% power: n=65 vs n=6500 for Canonne et al.Improved
Apply ToT framework (subsample-and-aggregate with private binomial test) at eps=1 with exact public test powergeneric public tests · Theoretical sample-size multiplier, alpha=0.05data multiplier upper boundNo clear effect
Tune m and alpha_0 for ToT with a z-test rather than using the upper bound (ablation)Test of Tests (z-test) · z-test alpha=0.05, effect size 0.65 sd, eps=1sample size for 80% power: n=70 vs n=100 boundImproved
Tune m and alpha_0 for ToT on ANOVA instead of Table 1 upper bound (ablation)Test of Tests (ANOVA, 3 groups) · eps=0.01, 95% powerdata multiplier: 41x vs 52x upper boundImproved
Use ToT rather than the Canonne et al. multivariate test (with tuned alpha matched to Type I error)multivariate normal mean test · d=100, eps=1, small effect mu_1=0.5sample size for 80% power: 190 vs 850Improved
Use ToT rather than Couch et al. nonparametric private ANOVA testone-way ANOVA, eta=1, eps=1, g=3 · ANOVA power comparisondata needed for 80% power: Couch et al. require 22% less dataHurt
Use ToT rather than Swanberg et al. ANOVA testone-way ANOVA, eta=1, eps=1, g=3 · ANOVA power comparisonpowerImproved
Use ToT rather than Couch et al. ANOVA test in smaller effect size settingone-way ANOVA · small effect size, g=3powerImproved
Use ToT rather than Couch et al. ANOVA test in large effect size and/or smaller eps settingsone-way ANOVA · large effect size or eps=0.1powerMixed
Use ToT rather than Canonne et al. multivariate test at 99% power (high power target)multivariate normal mean test, small effect · 99% powersample size needed: ToT needs 87% of Canonne dataImproved
Optimize ToT parameters m, alpha_0 with a high-m, high-alpha_0 setting rather than minimal m (PB's low-power recommendation) (ablation)Peña-Barrientos framework comparison · t-test eps=0.1, mu=3sample size for 80% power: n=125 vs n=348Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy