Open AccessWinning the NIST Contest: A scalable and general approach to differentially private synthetic data
Author(s) -
Ryan McKenna,
Gerome Miklau,
Daniel Sheldon
Publication year - 2021
Publication title -
the journal of privacy and confidentiality
Language(s) - English
Resource type - Journals
ISSN - 2575-8527
DOI - 10.29012/jpc.778
Subject(s) - nist , synthetic data , computer science , scalability , parametric statistics , contest , noise (video) , data mining , artificial intelligence , mathematics , statistics , speech recognition , database , political science , law , image (mathematics)
We propose a general approach for differentially private synthetic data generation, that consists of three steps: (1) select a collection of low-dimensional marginals, (2) measure those marginals with a noise addition mechanism, and (3) generate synthetic data that preserves the measured marginals well. Central to this approach is Private-PGM, a post-processing method that is used to estimate a high-dimensional data distribution from noisy measurements of its marginals. We present two mechanisms, NIST-MST and MST, that are instances of this general approach. NIST-MST was the winning mechanism in the 2018 NIST differential privacy synthetic data competition, and MST is a new mechanism that can work in more general settings, while still performing comparably to NIST-MST. We believe our general approach should be of broad interest, and can be adopted in future mechanisms for synthetic data generation.