A Central Limit Theorem for Differentially Private Query Answering

by   Jinshuo Dong, et al.

Perhaps the single most important use case for differential privacy is to privately answer numerical queries, which is usually achieved by adding noise to the answer vector. The central question, therefore, is to understand which noise distribution optimizes the privacy-accuracy trade-off, especially when the dimension of the answer vector is high. Accordingly, extensive literature has been dedicated to the question and the upper and lower bounds have been matched up to constant factors [BUV18, SU17]. In this paper, we take a novel approach to address this important optimality question. We first demonstrate an intriguing central limit theorem phenomenon in the high-dimensional regime. More precisely, we prove that a mechanism is approximately Gaussian Differentially Private [DRS21] if the added noise satisfies certain conditions. In particular, densities proportional to e^-x_p^α, where x_p is the standard ℓ_p-norm, satisfies the conditions. Taking this perspective, we make use of the Cramer–Rao inequality and show an "uncertainty principle"-style result: the product of the privacy parameter and the ℓ_2-loss of the mechanism is lower bounded by the dimension. Furthermore, the Gaussian mechanism achieves the constant-sharp optimal privacy-accuracy trade-off among all such noises. Our findings are corroborated by numerical experiments.


page 1

page 2

page 3

page 4


Convergence Rates for Differentially Private Statistical Estimation

Differential privacy is a cryptographically-motivated definition of priv...

Improved Utility Analysis of Private CountSketch

Sketching is an important tool for dealing with high-dimensional vectors...

Improved Differentially Private Euclidean Distance Approximation

This work shows how to privately and more accurately estimate Euclidean ...

An Uncertainty Principle is a Price of Privacy-Preserving Microdata

Privacy-protected microdata are often the desired output of a differenti...

The Discrete Gaussian for Differential Privacy

We show how to efficiently provide differentially private answers to cou...

Improving the Gaussian Mechanism for Differential Privacy: Analytical Calibration and Optimal Denoising

The Gaussian mechanism is an essential building block used in multitude ...

On The Differential Privacy of Thompson Sampling With Gaussian Prior

We show that Thompson Sampling with Gaussian Prior as detailed by Algori...