Minimax rate for multivariate data under componentwise local differential privacy constraints

TL;DR

This paper investigates the statistical limits of multivariate data analysis under componentwise local differential privacy, providing bounds and methods for density and covariance estimation, and analyzing information leakage risks.

Contribution

It introduces general techniques for minimax bounds under CLDP and applies them to density and covariance estimation, including adaptive procedures and privacy leakage analysis.

Findings

Established minimax bounds for density estimation under CLDP

Developed adaptive procedures matching bounds up to constants

Quantified privacy leakage risks between correlated components

Abstract

Our research delves into the balance between maintaining privacy and preserving statistical accuracy when dealing with multivariate data that is subject to \textit{componentwise local differential privacy} (CLDP). With CLDP, each component of the private data is made public through a separate privacy channel. This allows for varying levels of privacy protection for different components or for the privatization of each component by different entities, each with their own distinct privacy policies. We develop general techniques for establishing minimax bounds that shed light on the statistical cost of privacy in this context, as a function of the privacy levels $\alpha_1, ... , \alpha_d$ of the $d$ components. We demonstrate the versatility and efficiency of these techniques by presenting various statistical applications. Specifically, we examine nonparametric density and covariance estimation under CLDP, providing upper and lower bounds that match up to constant factors, as well as an associated data-driven adaptive procedure. Furthermore, we quantify the probability of extracting sensitive information from one component by exploiting the fact that, on another component which may be correlated with the first, a smaller degree of privacy protection is guaranteed.