Beyond One-Size-Fits-All: Neural Networks for Differentially Private Tabular Data Synthesis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Kai, Gong, Chen, Wang, Tianhao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917086547148800
author Chen, Kai
Gong, Chen
Wang, Tianhao
author_facet Chen, Kai
Gong, Chen
Wang, Tianhao
contents In differentially private (DP) tabular data synthesis, the consensus is that statistical models are better than neural network (NN)-based methods. However, we argue that this conclusion is incomplete and overlooks the challenge of densely correlated datasets, where intricate dependencies can overwhelm statistical models. In such complex scenarios, neural networks are more suitable due to their capacity to fit complex distributions by learning directly from samples. Despite this potential, existing NN-based algorithms still suffer from significant limitations. We therefore propose MargNet, incorporating successful algorithmic designs of statistical models into neural networks. MargNet applies an adaptive marginal selection strategy and trains the neural networks to generate data that conforms to the selected marginals. On sparsely correlated datasets, our approach achieves utility close to the best statistical method while offering an average 7$\times$ speedup over it. More importantly, on densely correlated datasets, MargNet establishes a new state-of-the-art, reducing fidelity error by up to 26\% compared to the previous best. We release our code on GitHub.\footnote{https://github.com/KaiChen9909/margnet}
format Preprint
id arxiv_https___arxiv_org_abs_2511_13893
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Beyond One-Size-Fits-All: Neural Networks for Differentially Private Tabular Data Synthesis
Chen, Kai
Gong, Chen
Wang, Tianhao
Machine Learning
Cryptography and Security
In differentially private (DP) tabular data synthesis, the consensus is that statistical models are better than neural network (NN)-based methods. However, we argue that this conclusion is incomplete and overlooks the challenge of densely correlated datasets, where intricate dependencies can overwhelm statistical models. In such complex scenarios, neural networks are more suitable due to their capacity to fit complex distributions by learning directly from samples. Despite this potential, existing NN-based algorithms still suffer from significant limitations. We therefore propose MargNet, incorporating successful algorithmic designs of statistical models into neural networks. MargNet applies an adaptive marginal selection strategy and trains the neural networks to generate data that conforms to the selected marginals. On sparsely correlated datasets, our approach achieves utility close to the best statistical method while offering an average 7$\times$ speedup over it. More importantly, on densely correlated datasets, MargNet establishes a new state-of-the-art, reducing fidelity error by up to 26\% compared to the previous best. We release our code on GitHub.\footnote{https://github.com/KaiChen9909/margnet}
title Beyond One-Size-Fits-All: Neural Networks for Differentially Private Tabular Data Synthesis
topic Machine Learning
Cryptography and Security
url https://arxiv.org/abs/2511.13893