Prior-Aware Distribution Estimation for Differential Privacy
This work addresses the challenge of accurate distribution estimation for privacy-focused applications like query answering and synthetic data generation, but it is incremental as it builds on existing methods with a specific optimization approach.
The paper tackles the problem of joint distribution estimation under differential privacy by using differentially private workload answers and a prior empirical distribution from public data, aiming to find a distribution that matches the workload accuracy while minimizing KL divergence with the prior, and their solution achieved second place in the NIST 2020 Differential Privacy Temporal Map Challenge.
Joint distribution estimation of a dataset under differential privacy is a fundamental problem for many privacy-focused applications, such as query answering, machine learning tasks and synthetic data generation. In this work, we examine the joint distribution estimation problem given two data points: 1) differentially private answers of a workload computed over private data and 2) a prior empirical distribution from a public dataset. Our goal is to find a new distribution such that estimating the workload using this distribution is as accurate as the differentially private answer, and the relative entropy, or KL divergence, of this distribution is minimized with respect to the prior distribution. We propose an approach based on iterative optimization for solving this problem. An application of our solution won second place in the NIST 2020 Differential Privacy Temporal Map Challenge, Sprint 2.