mlsanityData to 7 Oct 2026

Tabular Data Synthesis with Differential Privacy: A Survey

arXiv 2411.03351 ↗First posted 4 Nov 2024cs.CR cs.AI cs.DBcited by 6

Not on the map. Why?

Abstract (arXiv)

Data sharing is a prerequisite for collaborative innovation, enabling organizations to leverage diverse datasets for deeper insights. In real-world applications like FinTech and Smart Manufacturing, transactional data, often in tabular form, are generated and analyzed for insight generation. However, such datasets typically contain sensitive personal/business information, raising privacy concerns and regulatory risks. Data synthesis tackles this by generating artificial datasets that preserve the statistical characteristics of real data, removing direct links to individuals. However, attackers can still infer sensitive information using background knowledge. Differential privacy offers a solution by providing provable and quantifiable privacy protection. Consequently, differentially private data synthesis has emerged as a promising approach to privacy-aware data sharing. This paper provides a comprehensive overview of existing differentially private tabular data synthesis methods, highlighting the unique challenges of each generation model for generating tabular data under differential privacy constraints. We classify the methods into statistical and deep learning-based approaches based on their generation models, discussing them in both centralized and distributed environments. We evaluate and compare those methods within each category, highlighting their strengths and weaknesses in terms of utility, privacy, and computational complexity. Additionally, we present and discuss various evaluation methods for assessing the quality of the synthesized data, identify research gaps in the field and directions for future research.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 63 of its 142 references that are in the map. First posted on arXiv 4 Nov 2024.

Builds on

63 of 142 references.

References · 63

  1. 2022Generative Adversarial Networkscited by 5,603
  2. 2016Deep Learning with Differential Privacycited by 2,682
  3. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  4. 2014Conditional Generative Adversarial Netscited by 2,152
  5. 2019Modeling Tabular data using Conditional GANcited by 695
  6. 2017Renyi Differential Privacycited by 643
  7. 2022TabDDPM: Modelling Tabular Data with Diffusion Modelscited by 373
  8. 2016Semi-supervised Knowledge Transfer for Deep Learning from Private Training Datacited by 347
  9. 2016Concentrated Differential Privacy: Simplifications, Extensions, and Lower Boundscited by 339
  10. 2013The Composition Theorem for Differential Privacycited by 290
  11. 2022Language Models are Realistic Tabular Data Generatorscited by 290
  12. 2018Differentially Private Generative Adversarial Networkcited by 235
38 more
  1. 2018Improving the Gaussian Mechanism for Differential Privacy: Analytical Calibration and Optimal Denoisingcited by 216
  2. 2017Generating Multi-label Discrete Patient Records using Generative Adversarial Networkscited by 197
  3. 2021How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Modelscited by 193
  4. 2024Large Language Models(LLMs) on Tabular Data: Prediction, Generation, and Understanding -- A Surveycited by 157
  5. 2021CTAB-GAN: Effective Table Data Synthesizingcited by 156
  6. 2023Machine Learning for Synthetic Data Generation: A Reviewcited by 118
  7. 2020Synthetic Data -- Anonymisation Groundhog Daycited by 105
  8. 2021Winning the NIST Contest: A scalable and general approach to differentially private synthetic datacited by 101
  9. 2022AIM: An Adaptive and Iterative Mechanism for Differentially Private Synthetic Datacited by 88
  10. 2018An empirical study on evaluation metrics of generative adversarial networkscited by 81
  11. 2020DP-MERF: Differentially Private Mean Embeddings with Random Features for Practical Privacy-Preserving Data Generationcited by 78
  12. 2023CoDi: Co-evolving Contrastive Diffusion Models for Mixed-type Tabular Synthesiscited by 78
  13. 2022Synthetic Data -- what, why and how?cited by 66
  14. 2023Generating and Imputing Tabular Data via Diffusion and Flow-based Gradient-Boosted Treescited by 60
  15. 2024Comprehensive Exploration of Synthetic Data Generation: A Surveycited by 59
  16. 2021Iterative Methods for Private Synthetic Data: Unifying Framework and New Methodscited by 58
  17. 2019G-PATE: Scalable Differentially Private Data Generator via Private Aggregation of Teacher Discriminatorscited by 53
  18. 2020Differentially Private Synthetic Medical Data Generation using Convolutional GANscited by 52
  19. 2020Variational Autoencoders for Learning Nonlinear Dynamics of Physical Systemscited by 50
  20. 2021Differentially Private Query Release Through Adaptive Projectioncited by 50
  21. 2021Robin Hood and Matthew Effects: Differential Privacy Has Disparate Impact on Synthetic Datacited by 48
  22. 2022A Unified Framework for Quantifying Privacy Risk in Synthetic Datacited by 45
  23. 2020New Oracle-Efficient Algorithms for Private Synthetic Data Releasecited by 43
  24. 2023FinDiff: Diffusion Models for Financial Tabular Data Generationcited by 41
  25. 2017Differentially Private Mixture of Generative Neural Networkscited by 39
  26. 2022Private Synthetic Data for Multitask Learning and Marginal Queriescited by 35
  27. 2018Differentially Private Data Generative Modelscited by 33
  28. 2019Differentially Private Generative Adversarial Networks for Time Series, Continuous, and Discrete Open Datacited by 31
  29. 2014Dual Query: Practical Private Query Release for High Dimensional Datacited by 26
  30. 2023AutoDiff: combining Auto-encoder and Diffusion model for tabular data synthesizingcited by 24
  31. 2021PEARL: Data Synthesis via Private Embeddings and Adversarial Reconstruction Learningcited by 23
  32. 2023A Linear Reconstruction Approach for Attribute Inference Attacks against Synthetic Datacited by 23
  33. 2020P3GM: Private High-Dimensional Data Release via Privacy Preserving Phased Generative Modelcited by 21
  34. 2020Relational Data Synthesis using Generative Adversarial Networks: A Design Space Explorationcited by 21
  35. 2021Hermite Polynomial Features for Private Data Generationcited by 21
  36. 2020Bias and Variance of Post-processing in Differential Privacycited by 14
  37. 2016Comparative Study of Differentially Private Data Synthesis Methodscited by 13
  38. 2024Balanced Mixed-Type Tabular Data Synthesis with Diffusion Modelscited by 12

50 most cited of 63 All 63 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 4; 2026: 2.

6 papers.

  1. 2025Is API Access to LLMs Useful for Generating Private Synthetic Tabular Data?cited by 5
  2. 2026Differentially Private Synthetic Data via APIs 4: Tabular Datacited by 2
  3. 2025Privacy-Preserving Dataset Combinationcited by 1
  4. 2025PrAda-GAN: A Private Adaptive Generative Adversarial Network with Bayes Network Structurecited by 1
  5. 2025Measuring LLM Sensitivity in Transformer-based Tabular Data Synthesiscited by 0
  6. 2026Measuring Privacy Risks and Tradeoffs in Financial Synthetic Data Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Differentially Private Tabular Data Synthesis using Large Language Models
  2. 2024Systematic Assessment of Tabular Data Synthesis
  3. 2024Federated Diffusion Modeling with Differential Privacy for Tabular Data Synthesis
  4. 2026Creation begins with understanding: LLMs as strategy designers for privacy-preserving tabular data synthesis
  5. 2025Membership Inference over Diffusion-models-based Synthetic Tabular Data
  6. 2023Generating tabular datasets under differential privacy
  7. 2024HARMONIC: Harnessing LLMs for Tabular Data Synthesis and Privacy Protection
  8. 2025Flow Matching for Tabular Data Synthesis