Timezone: »
Canonical Noise Distributions and Private Hypothesis Tests
Jordan Awan · Salil Vadhan
Event URL: https://openreview.net/forum?id=pr_6JwMBLa1 »
In the setting of $f$-DP, we propose the concept \emph{canonical noise distribution} (CND) which captures whether an additive privacy mechanism is tailored for a given $f$, and give a construction of a CND for an arbitrary tradeoff function $f$. We show that private hypothesis tests are intimately related to CNDs, allowing for the release of private $p$-values at no additional privacy cost as well as the construction of uniformly most powerful (UMP) tests for binary data. We apply our techniques to difference of proportions testing.
In the setting of $f$-DP, we propose the concept \emph{canonical noise distribution} (CND) which captures whether an additive privacy mechanism is tailored for a given $f$, and give a construction of a CND for an arbitrary tradeoff function $f$. We show that private hypothesis tests are intimately related to CNDs, allowing for the release of private $p$-values at no additional privacy cost as well as the construction of uniformly most powerful (UMP) tests for binary data. We apply our techniques to difference of proportions testing.
Author Information
Jordan Awan (Penn State University)
Salil Vadhan (Harvard University)
Related Events (a corresponding poster, oral, or spotlight)
-
2021 : Canonical Noise Distributions and Private Hypothesis Tests »
Dates n/a. Room
More from the Same Authors
-
2021 : Privacy-Aware Rejection Sampling »
Jordan Awan · Vinayak Rao -
2021 : Privacy-Aware Rejection Sampling »
Jordan Awan · Vinayak Rao -
2019 Poster: Elliptical Perturbations for Differential Privacy »
Matthew Reimherr · Jordan Awan -
2019 Poster: KNG: The K-Norm Gradient Mechanism »
Matthew Reimherr · Jordan Awan -
2018 Poster: Differentially Private Uniformly Most Powerful Tests for Binomial Data »
Jordan Awan · Aleksandra Slavković -
2016 Poster: Privacy Odometers and Filters: Pay-as-you-Go Composition »
Ryan Rogers · Salil Vadhan · Aaron Roth · Jonathan Ullman