Winning the NIST Contest: A scalable and general approach to differentially private synthetic data
read the original abstract
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.
This paper has not been read by Pith yet.
Forward citations
Cited by 7 Pith papers
-
SoK: Reconstruction Attacks on Synthetic Tabular Data (Insights from Winning the NIST CRC)
The first systematization of reconstruction attacks on synthetic tabular data finds that generator choice dominates privacy risk over attack choice, with differential privacy effective only at low budgets and most lea...
-
DPDSyn: Improving Differentially Private Dataset Synthesis for Model Training by Downstream Task Guidance
DPDSyn guides differentially private dataset synthesis with a downstream-task model trained under DP, yielding higher accuracy and faster generation than prior distribution-selection methods.
-
A Dual Perspective on Synthetic Trajectory Generators: Utility Framework and Privacy Vulnerabilities
A new framework evaluates utility of synthetic mobility trajectories while a membership inference attack reveals privacy vulnerabilities in generative models thought to be safe.
-
Quantifying the Privacy of Counterfactuals by Leveraging Membership Inference Attacks Against Synthetic Data
Membership inference attacks adapted from synthetic data succeed on counterfactuals using only the counterfactuals themselves, without model access.
-
Measuring Database Unfairness via Dependency Quantification Under Differential Privacy
Proposes a formal DP-compatible framework with three unfairness measures (mutual information with TV proxy, MaxSAT-based repair, top-k tuple contribution) that satisfy positivity, monotonicity, and computability.
-
Aim High, Stay Private: Differentially Private Synthetic Data Enables Public Release of Behavioral Health Information with High Utility
The authors apply the Adaptive Iterative Mechanism to create differentially private synthetic data from the LEMURS wearable and survey dataset and show that epsilon=5 retains useful predictive performance for downstre...
-
Decoupling Identity from Utility: Privacy-by-Design Frameworks for Financial Ecosystems
Differentially private synthetic data and seeded agent-based models can separate personal identities from usable financial data while meeting regulatory privacy rules.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.