Correlated Noise Mechanisms for Differentially Private Learning

TL;DR

This paper investigates correlated noise mechanisms for differential privacy in machine learning, showing how introducing correlations can enhance privacy-utility trade-offs and are practically deployed at scale.

Contribution

It analyzes the design and analysis of correlated noise mechanisms like matrix and factorization mechanisms for DP, highlighting their advantages over independent noise.

Findings

Correlated noise mechanisms improve privacy-utility trade-offs.

These mechanisms are effectively used in large-scale industrial applications.

Correlations in noise can cancel out some noise effects in iterative algorithms.

Abstract

This monograph explores the design and analysis of correlated noise mechanisms for differential privacy (DP), focusing on their application to private training of AI and machine learning models via the core primitive of estimation of weighted prefix sums. While typical DP mechanisms inject independent noise into each step of a stochastic gradient (SGD) learning algorithm in order to protect the privacy of the training data, a growing body of recent research demonstrates that introducing (anti-)correlations in the noise can significantly improve privacy-utility trade-offs by carefully canceling out some of the noise added on earlier steps in subsequent steps. Such correlated noise mechanisms, known variously as matrix mechanisms, factorization mechanisms, and DP-Follow-the-Regularized-Leader (DP-FTRL) when applied to learning algorithms, have also been influential in practice, with industrial deployment at a global scale.