One Step to Efficient Synthetic Data

Abstract

We propose a general method of producing synthetic data, which is widely applicable for parametric models, has asymptotically efficient summary statistics, and is both easily implemented and highly computationally efficient. Our approach allows for the construction of both partially synthetic datasets, which preserve the summary statistics without formal privacy methods, as well as fully synthetic data which satisfy the strong guarantee of differential privacy (DP), both with asymptotically efficient summary statistics. While our theory deals with asymptotics, we demonstrate through simulations that our approach offers high utility in small samples as well. In particular we 1) apply our method to the Burr distribution, evaluating the parameter estimates as well as distributional properties with the Kolmogorov-Smirnov test, 2) demonstrate the performance of our mechanism on a log-linear model based on a car accident dataset, and 3) produce DP synthetic data for the beta distribution using a customized Laplace mechanism.Preprint. Under review.