Differentially Private Synthetic Data via APIs 4: Tabular Data

Toan Tran, Arturs Backurs, Zinan Lin, Victor Reis, Li Xiong, Sergey Yekhanin
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:121952-121984, 2026.

Abstract

This paper investigates the problem of generating synthetic tabular data with differential privacy (DP) guarantees, enabling data sharing in sensitive domains. Despite extensive study, state-of-the-art methods often focus on minimizing low-order marginal query errors and overlook the challenges posed by high-order correlations. To address this gap, we extend the Private Evolution (PE) framework, originally developed for DP-compliant image and text synthesis, to tabular data. We introduce Tab-PE – an algorithm for synthetic tabular data generation under DP constraints. Tab-PE iteratively improves a candidate dataset via an evolutionary process that leverages tabular-specialized operators to produce variations, privately scores them, and selects the highest-quality samples to retain and propagate. In contrast to the original PE, which relies on large foundation models, Tab-PE employs heuristic operators with significantly lower computational costs, making PE more practical and scalable for tabular data. Through extensive experiments on real-world and simulation datasets, we demonstrate that Tab-PE substantially outperforms prior baselines on datasets exhibiting high-order correlations. Compared to the best baseline – AIM, Tab-PE improves classification accuracy by up to 10% while running 28$\times$ faster.

Cite this Paper


BibTeX
@InProceedings{pmlr-v306-tran26a, title = {Differentially Private Synthetic Data via {API}s 4: Tabular Data}, author = {Tran, Toan and Backurs, Arturs and Lin, Zinan and Reis, Victor and Xiong, Li and Yekhanin, Sergey}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {121952--121984}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/tran26a/tran26a.pdf}, url = {https://proceedings.mlr.press/v306/tran26a.html}, abstract = {This paper investigates the problem of generating synthetic tabular data with differential privacy (DP) guarantees, enabling data sharing in sensitive domains. Despite extensive study, state-of-the-art methods often focus on minimizing low-order marginal query errors and overlook the challenges posed by high-order correlations. To address this gap, we extend the Private Evolution (PE) framework, originally developed for DP-compliant image and text synthesis, to tabular data. We introduce Tab-PE – an algorithm for synthetic tabular data generation under DP constraints. Tab-PE iteratively improves a candidate dataset via an evolutionary process that leverages tabular-specialized operators to produce variations, privately scores them, and selects the highest-quality samples to retain and propagate. In contrast to the original PE, which relies on large foundation models, Tab-PE employs heuristic operators with significantly lower computational costs, making PE more practical and scalable for tabular data. Through extensive experiments on real-world and simulation datasets, we demonstrate that Tab-PE substantially outperforms prior baselines on datasets exhibiting high-order correlations. Compared to the best baseline – AIM, Tab-PE improves classification accuracy by up to 10% while running 28$\times$ faster.} }
Endnote
%0 Conference Paper %T Differentially Private Synthetic Data via APIs 4: Tabular Data %A Toan Tran %A Arturs Backurs %A Zinan Lin %A Victor Reis %A Li Xiong %A Sergey Yekhanin %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-tran26a %I PMLR %P 121952--121984 %U https://proceedings.mlr.press/v306/tran26a.html %V 306 %X This paper investigates the problem of generating synthetic tabular data with differential privacy (DP) guarantees, enabling data sharing in sensitive domains. Despite extensive study, state-of-the-art methods often focus on minimizing low-order marginal query errors and overlook the challenges posed by high-order correlations. To address this gap, we extend the Private Evolution (PE) framework, originally developed for DP-compliant image and text synthesis, to tabular data. We introduce Tab-PE – an algorithm for synthetic tabular data generation under DP constraints. Tab-PE iteratively improves a candidate dataset via an evolutionary process that leverages tabular-specialized operators to produce variations, privately scores them, and selects the highest-quality samples to retain and propagate. In contrast to the original PE, which relies on large foundation models, Tab-PE employs heuristic operators with significantly lower computational costs, making PE more practical and scalable for tabular data. Through extensive experiments on real-world and simulation datasets, we demonstrate that Tab-PE substantially outperforms prior baselines on datasets exhibiting high-order correlations. Compared to the best baseline – AIM, Tab-PE improves classification accuracy by up to 10% while running 28$\times$ faster.
APA
Tran, T., Backurs, A., Lin, Z., Reis, V., Xiong, L. & Yekhanin, S.. (2026). Differentially Private Synthetic Data via APIs 4: Tabular Data. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:121952-121984 Available from https://proceedings.mlr.press/v306/tran26a.html.

Related Material