dpmm: Differentially Private Marginal Models, a Library for Synthetic Tabular Data Generation

TL;DR

dpmm is an open-source library that generates synthetic tabular data with differential privacy guarantees, integrating popular marginal models and best practices for robustness and utility.

Contribution

It introduces a comprehensive, customizable library combining multiple marginal models with end-to-end differential privacy guarantees for synthetic data generation.

Findings

Achieves superior utility compared to existing methods

Provides robust, customizable implementations of PrivBayes, MST, and AIM

Addresses known DP vulnerabilities effectively

Abstract

We propose dpmm, an open-source library for synthetic data generation with Differentially Private (DP) guarantees. It includes three popular marginal models -- PrivBayes, MST, and AIM -- that achieve superior utility and offer richer functionality compared to alternative implementations. Additionally, we adopt best practices to provide end-to-end DP guarantees and address well-known DP-related vulnerabilities. Our goal is to accommodate a wide audience with easy-to-install, highly customizable, and robust model implementations. Our codebase is available from https://github.com/sassoftware/dpmm.