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

Abstract

We present a generic framework for creating differentially private versions of any hypothesis test in a black-box way. We analyze the resulting tests analytically and experimentally. Most crucially, we show good practical performance for small data sets, showing that at ε = 1 we only need 5-6 times as much data as in the fully public setting. We compare our work to the one existing framework of this type, as well as to several individually-designed private hypothesis tests. Our framework is higher power than other generic solutions and at least competitive with (and often better than) individually-designed tests.

Cite

Text

Kazan et al. "The Test of Tests: A Framework for Differentially Private Hypothesis Testing." International Conference on Machine Learning, 2023.

Markdown

[Kazan et al. "The Test of Tests: A Framework for Differentially Private Hypothesis Testing." International Conference on Machine Learning, 2023.](https://mlanthology.org/icml/2023/kazan2023icml-test/)

BibTeX

@inproceedings{kazan2023icml-test,
  title     = {{The Test of Tests: A Framework for Differentially Private Hypothesis Testing}},
  author    = {Kazan, Zeki and Shi, Kaiyan and Groce, Adam and Bray, Andrew P},
  booktitle = {International Conference on Machine Learning},
  year      = {2023},
  pages     = {16131-16151},
  volume    = {202},
  url       = {https://mlanthology.org/icml/2023/kazan2023icml-test/}
}