English

Joint Selection: Adaptively Incorporating Public Information for Private Synthetic Data

Machine Learning 2024-03-13 v1 Artificial Intelligence

Abstract

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.

Keywords

Cite

@article{arxiv.2403.07797,
  title  = {Joint Selection: Adaptively Incorporating Public Information for Private Synthetic Data},
  author = {Miguel Fuentes and Brett Mullins and Ryan McKenna and Gerome Miklau and Daniel Sheldon},
  journal= {arXiv preprint arXiv:2403.07797},
  year   = {2024}
}
R2 v1 2026-06-28T15:17:31.957Z