Fidelity-agnostic synthetic data generation improves utility while retaining privacy
The DOI matches its Crossref record; 4 of the 4 references it cites were found in a registry or an index.
Abstract
Synthetic data are a popular method to publish useful datasets in a privacy-aware manner, making them useful across a range of scientific domains involving human subjects. They are typically generated by sampling from algorithms that mimic the probability distribution of real datasets, thereby maximizing statistical similarity to real data. However, we argue and demonstrate that synthetic data need to be similar only in ways relevant to their intended use and may neglect any irrelevant information, which in turn may improve privacy protection. As such, we propose a data synthesis method entitled fidelity-agnostic synthetic data. The method first extracts features relevant to the dataset's intended use using a neural net and then generates synthetic versions of the extracted features, after which they are decoded to mimic the real dataset. We show that our synthetic data improve performance in prediction tasks while retaining privacy protection compared to other state-of-the-art methods.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Denoising Diffusion Probabilistic Models
Ho, J., Jain, A., & Abbeel, P. (2020) · arXiv · doi:10.48550/arxiv.2006.11239 · Cited by 8,712 · cited 1×
“We opt for a VAE, but diffusion models, 61 for example, would be an equally valid choice.”
Yeh, I.-C., & Lien, C.-H. (2009) · Expert Systems with Applications · doi:10.1016/j.eswa.2007.12.020 · Cited by 885
Proceedings of the ACM Web Conference 2022
(2022) · doi:10.1145/3485447 · Cited by 522
Synthetic data generation for tabular health records: A systematic review
Hernandez, M., Epelde, G., Alberdi, A., Cilla, R., & Rankin, D. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.04.053 · Cited by 278
Research works printed but not linked: 32
- van Kesteren, 2024, To democratize research with sensitive data, we should make synthetic data more accessible, Patterns, 5, 10.1016/j.patter.2024.101049 · doi:10.1016/j.patter.2024.101049
- Jordon, 2022, Synthetic data–what, why and how?, arXiv
- van Dijk, 2024, A novel taxonomy for navigating and classifying synthetic data in healthcare applications, Stud. Health Technol. Inform., 321, 259
- Goetz, 2020, Federated learning via synthetic data, arXiv
- Yan, 2022, A multifaceted benchmarking of synthetic electronic health record generation models, Nat. Commun., 13, 7609, 10.1038/s41467-022-35295-1 · DOI · Paper page
- Liu, 2021, Iterative methods for private synthetic data: Unifying framework and new methods, Adv. Neural Inf. Process. Syst., 34, 690
- Chen, 2022, Private set generation with discriminative information, Adv. Neural Inf. Process. Syst., 35, 14678
- Bengio, 2013, Representation learning: A review and new perspectives, IEEE Trans. Pattern Anal. Mach. Intell., 35, 1798, 10.1109/TPAMI.2013.50 · DOI · Paper page
- Achterberg, 2024, On the evaluation of synthetic longitudinal electronic health records, BMC Med. Res. Methodol., 24, 181, 10.1186/s12874-024-02304-4 · DOI
- El Emam, 2008, Protecting privacy using k-anonymity, J. Am. Med. Inform. Assoc., 15, 627, 10.1197/jamia.M2716 · DOI · Paper page
- Yoon, 2020, Anonymization through data synthesis using generative adversarial networks (ads-gan), IEEE J. Biomed. Health Inform., 24, 2378, 10.1109/JBHI.2020.2980262 · DOI · Paper page
- Rodriguez-Almeida, 2023, Synthetic patient data generation and evaluation in disease prediction using small and imbalanced datasets, IEEE J. Biomed. Health Inform., 27, 2670, 10.1109/JBHI.2022.3196697 · DOI
- Khan, 2024, Utility of gan generated synthetic data for cardiovascular diseases mortality prediction: an experimental study, Health Technol., 14, 557, 10.1007/s12553-024-00847-6 · DOI
- Sun, 2019, Learning vine copula models for synthetic data generation, Proc. AAAI Conf. Artif. Intell., 33, 5049
- Esteban, 2017, Real-valued (medical) time series generation with recurrent conditional gans, arXiv
- Xie, 2018, Differentially private generative adversarial network, arXiv
- Agarwal, C., Krishna, S., Saxena, E., Pawelczyk, M., Johnson, N., Puri, I., Zitnik, M., and Lakkaraju, H. (2022). Openxai: Towards a transparent evaluation of model explanations. In: Advances in Neural Information Processing Systems vol. 35. ( 15784–15799). URL: https://proceedings.neurips.cc/paper_files/paper/2022/file/65398a0eba88c9b4a1c38ae405b125ef-Paper-Datasets_and_Benchmarks.pdf.
- Zhang, 2017, Privbayes: Private data release via bayesian networks, ACM Trans. Database Syst., 42, 1, 10.1145/3134428 · DOI · Paper page
- Mukherjee, 2021, Smote-enc: A novel smote-based method to generate synthetic data for nominal and continuous features, Appl. Syst. Innov., 4, 18, 10.3390/asi4010018 · DOI · Paper page
- Detrano, 1989, International application of a new probability algorithm for the diagnosis of coronary artery disease, Am. J. Cardiol., 64, 304, 10.1016/0002-9149(89)90524-9 · DOI · Paper page
- Kaur, 2021, Application of bayesian networks to generate synthetic health data, J. Am. Med. Inform. Assoc., 28, 801, 10.1093/jamia/ocaa303 · DOI
- Holm, 1979, A simple sequentially rejective multiple test procedure, Scand. J. Stat., 6, 65
- Ma, C., Tschiatschek, S., Turner, R., Hernández-Lobato, J. M., and Zhang, C. (2020). Vaem: a deep generative model for heterogeneous mixed type data. In: Advances in Neural Information Processing Systems vol. 33. (11237–11247). URL: https://proceedings.neurips.cc/paper_files/paper/2020/file/8171ac2c5544a5cb54ac0f38bf477af4-Paper.pdf.
- Gallos, 2024, Insafedare project: Innovative applications of assessment and assurance of data and synthetic data for regulatory decision support, Stud. Health Technol. Inform., 316, 1193
- Langevin, 2022, Generative adversarial networks for data augmentation and transfer in credit card fraud detection, J. Oper. Res. Soc., 73, 153, 10.1080/01605682.2021.1880296 · DOI
- Ucar, 2021, Subtab: Subsetting features of tabular data for self-supervised representation learning, Adv. Neural Inf. Process. Syst., 34, 18853
- Creswell, 2018, Generative adversarial networks: An overview, IEEE Signal Process. Mag., 35, 53, 10.1109/MSP.2017.2765202 · DOI · Paper page
- Torfi, 2022, Differentially private synthetic medical data generation using convolutional gans, Inf. Sci., 586, 485, 10.1016/j.ins.2021.12.018 · DOI
- Qian, 2024, Synthetic data for privacy-preserving clinical risk prediction, Sci. Rep., 14, 10.1038/s41598-024-72894-y · DOI
- Papernot, 2020, Making the shoe fit: Architectures, initializations, and tuning for learning with privacy, OpenReview
- Ward, 2024, Data plagiarism index: Characterizing the privacy risk of data-copying in tabular generative models, arXiv
- Achterberg, J., Haas, M., Van Dijk, B., and Spruit, M. (2025). Fidelity agnostic synthetic data generation improves utility whilst retaining privacy (repository). Zenodo. https://doi.org/10.5281/zenodo.15000502. · doi:10.21203/rs.3.rs-5319537/v1
Metadata
- Authors
- Achterberg, J., Haas, M., van Dijk, B., & Spruit, M.
- Venue
- Patterns
- Publisher
- Elsevier BV
- Published
- 2025-10
- Volume
- 6
- Issue
- 10
- Pages
- 101287
- ISSN
- 2666-3899
- Cited by
- 8
- References
- 4
- Language
- en
- Licence
- www.elsevier.com/tdm/userlicense/1.0/ (tdm)
- www.elsevier.com/legal/tdmrep-license (tdm)
- creativecommons.org/licenses/by/4.0/ (vor)
- Funders
- European Commission
- HORIZON EUROPE Health (101095661)
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Fidelity-agnostic synthetic data generation improves utility while retaining privacy. https://app.cochat.ai/papers/10.1016/j.patter.2025.101287
- The paper
- Achterberg, J., Haas, M., van Dijk, B., & Spruit, M. (2025). Fidelity-agnostic synthetic data generation improves utility while retaining privacy. Patterns. https://doi.org/10.1016/j.patter.2025.101287
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.