Joint Selection: Adaptively Incorporating Public Information for Private Synthetic Data
This addresses a limitation in synthetic data generation for privacy-preserving applications, though it appears incremental as it builds on adaptive measurements and graphical models.
The paper tackles the problem of generating differentially private synthetic data by incorporating public data, which existing methods cannot do when model structure is unknown, and shows that their mechanism jam-pgm outperforms both publicly assisted and non-publicly assisted methods even with biased public data.
Mechanisms for generating differentially private synthetic data based on marginals and graphical models have been successful in a wide range of settings. However, one limitation of these methods is their inability to incorporate public data. Initializing a data generating model by pre-training on public data has shown to improve the quality of synthetic data, but this technique is not applicable when model structure is not determined a priori. We develop the mechanism jam-pgm, which expands the adaptive measurements framework to jointly select between measuring public data and private data. This technique allows for public data to be included in a graphical-model-based mechanism. We show that jam-pgm is able to outperform both publicly assisted and non publicly assisted synthetic data generation mechanisms even when the public data distribution is biased.