Accurate and Scalable Matrix Mechanisms via Divide and Conquer

TL;DR

This paper introduces QuerySmasher, a divide-and-conquer matrix mechanism that improves scalability and accuracy for differentially private query answering by decomposing workloads into manageable sub-problems.

Contribution

QuerySmasher offers a novel scalable approach that generalizes and outperforms prior matrix mechanisms like ResidualPlanner and Weighted Fourier Factorizations.

Findings

QuerySmasher subsumes prior mechanisms and can dominate them under sum squared error.

Experimental results show QuerySmasher's scalability and improved accuracy.

The method effectively decomposes complex workloads into low-dimensional sub-problems.

Abstract

Matrix mechanisms are often used to provide unbiased differentially private query answers when publishing statistics or creating synthetic data. Recent work has developed matrix mechanisms, such as ResidualPlanner and Weighted Fourier Factorizations, that scale to high dimensional datasets while providing optimality guarantees for workloads such as marginals and circular product queries. They operate by adding noise to a linearly independent set of queries that can compactly represent the desired workloads. In this paper, we present QuerySmasher, an alternative scalable approach based on a divide-and-conquer strategy. Given a workload that can be answered from various data marginals, QuerySmasher splits each query into sub-queries and re-assembles the pieces into mutually orthogonal sub-workloads. These sub-workloads represent small, low-dimensional problems that can be independently and optimally answered by existing low-dimensional matrix mechanisms. QuerySmasher then stitches these solutions together to answer queries in the original workload. We show that QuerySmasher subsumes prior work, like ResidualPlanner (RP), ResidualPlanner+ (RP+), and Weighted Fourier Factorizations (WFF). We prove that it can dominate those approaches, under sum squared error, for all workloads. We also experimentally demonstrate the scalability and accuracy of QuerySmasher.