Research map / Federated learning and privacy / Differential privacy
Differentially Private Identity and Equivalence Testing of Discrete Distributions
Maryam Aliakbarpour, Ilias Diakonikolas, Ronitt Rubinfeld
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 |
|---|---|---|---|
| Use unique-elements statistic with Laplace noise (Private-Unique-Elements-Uniformity) instead of collisions-based tester | uniformity testing · n=800,000, eps=0.3, xi=0.2 synthetic | empirical error probability | Improved |
| Use unique-elements tester when sample size exceeds domain size (pn/eps^2 comparable to n) | uniformity testing · n=1000, eps=0.1 | usable guarantees | Hurt |
| Private identity tester built from reduction to uniformity with unique-elements tester | identity testing · n=800,000, eps=0.3, xi=0.2, uniform null | max of type I and type II error | Improved |
| Private equivalence testing via adding Laplace noise to chi-square statistic | equivalence testing · n from 1e4 to 2e6, eps=0.3, xi=0.2 | sample complexity | No clear effect |
Closest papers by meaning
- Minimax Optimal Two-Sample Testing under Local Differential Privacy (NeurIPS 2026 · similarity 0.90)
- Differentially Private Equivalence Testing for Continuous Distributions and Applications (NeurIPS 2024 · similarity 0.90)
- Sort, Partition, Randomize: Optimal Binary Hypothesis Testing under Local Differential Privacy (NeurIPS 2026 · similarity 0.89)
- Replicable Distribution Testing (NeurIPS 2025 · similarity 0.87)
- Optimal Private and Communication Constraint Distributed Goodness-of-Fit Testing for Discrete Distributions in the Large Sample Regime (NeurIPS 2024 · similarity 0.87)
- Optimal Rates for Differentially Private Hypothesis Testing with E-values (NeurIPS 2026 · 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.
