{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,15]],"date-time":"2025-11-15T10:34:41Z","timestamp":1763202881039,"version":"3.37.3"},"reference-count":43,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2024,6,14]],"date-time":"2024-06-14T00:00:00Z","timestamp":1718323200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2024,6,14]],"date-time":"2024-06-14T00:00:00Z","timestamp":1718323200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100002341","name":"Research Council of Finland","doi-asserted-by":"crossref","award":["325572","325572"],"award-info":[{"award-number":["325572","325572"]}],"id":[{"id":"10.13039\/501100002341","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100002341","name":"Research Council of Finland","doi-asserted-by":"crossref","award":["325573","325572"],"award-info":[{"award-number":["325573","325572"]}],"id":[{"id":"10.13039\/501100002341","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100000780","name":"European Union","doi-asserted-by":"crossref","award":["Project 101070617"],"award-info":[{"award-number":["Project 101070617"]}],"id":[{"id":"10.13039\/501100000780","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Strategic Research Council (SRC) established within the Research Council of Finland","award":["336032"],"award-info":[{"award-number":["336032"]}]},{"DOI":"10.13039\/100014013","name":"UK Research and Innovation","doi-asserted-by":"publisher","award":["EP\/W002973\/1"],"award-info":[{"award-number":["EP\/W002973\/1"]}],"id":[{"id":"10.13039\/100014013","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["BMC Med Inform Decis Mak"],"abstract":"<jats:title>Abstract<\/jats:title><jats:sec>\n                <jats:title>Background<\/jats:title>\n                <jats:p>Consider a setting where multiple parties holding sensitive data aim to collaboratively learn population level statistics, but pooling the sensitive data sets is not possible due to privacy concerns and parties are unable to engage in centrally coordinated joint computation. We study the feasibility of combining privacy preserving synthetic data sets in place of the original data for collaborative learning on real-world health data from the UK Biobank.<\/jats:p>\n              <\/jats:sec><jats:sec>\n                <jats:title>Methods<\/jats:title>\n                <jats:p>We perform an empirical evaluation based on an existing prospective cohort study from the literature. Multiple parties were simulated by splitting the UK Biobank cohort along assessment centers, for which we generate synthetic data using differentially private generative modelling techniques. We then apply the original study\u2019s Poisson regression analysis on the combined synthetic data sets and evaluate the effects of 1) the size of local data set, 2) the number of participating parties, and 3) local shifts in distributions, on the obtained likelihood scores.<\/jats:p>\n              <\/jats:sec><jats:sec>\n                <jats:title>Results<\/jats:title>\n                <jats:p>We discover that parties engaging in the collaborative learning via shared synthetic data obtain more accurate estimates of the regression parameters compared to using only their local data. This finding extends to the difficult case of small heterogeneous data sets. Furthermore, the more parties participate, the larger and more consistent the improvements become up to a certain limit. Finally, we find that data sharing can especially help parties whose data contain underrepresented groups to perform better-adjusted analysis for said groups.<\/jats:p>\n              <\/jats:sec><jats:sec>\n                <jats:title>Conclusions<\/jats:title>\n                <jats:p>Based on our results we conclude that sharing of synthetic data is a viable method for enabling learning from sensitive data without violating privacy constraints even if individual data sets are small or do not represent the overall population well. Lack of access to distributed sensitive data is often a bottleneck in biomedical research, which our study shows can be alleviated with privacy-preserving collaborative learning methods.<\/jats:p>\n              <\/jats:sec>","DOI":"10.1186\/s12911-024-02563-7","type":"journal-article","created":{"date-parts":[[2024,6,14]],"date-time":"2024-06-14T16:02:17Z","timestamp":1718380937000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":3,"title":["Collaborative learning from distributed data with differentially private synthetic data"],"prefix":"10.1186","volume":"24","author":[{"given":"Lukas","family":"Prediger","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Joonas","family":"J\u00e4lk\u00f6","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Antti","family":"Honkela","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Samuel","family":"Kaski","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"297","published-online":{"date-parts":[[2024,6,14]]},"reference":[{"key":"2563_CR1","doi-asserted-by":"crossref","unstructured":"Dwork C, McSherry F, Nissim K, Smith AD. Calibrating Noise to Sensitivity in Private Data Analysis. In: 3rd Theory of Cryptography Conf. Berlin, Heidelberg: Springer; 2006. p. 265\u201384.","DOI":"10.1007\/11681878_14"},{"key":"2563_CR2","unstructured":"Hardt M, Ligett K, McSherry F. A Simple and Practical Algorithm for Differentially Private Data Release. In: Adv. Neural Inf. Process. Syst. Red Hook: Curran Associates, Inc; 2012. p. 2339\u201347."},{"key":"2563_CR3","doi-asserted-by":"crossref","unstructured":"Chen R, Acs G, Castelluccia C. Differentially Private Sequential Data Publication via Variable-length n-grams. In: Proc. 2012 ACM Conf. Comput. and Commun. Security. New York: ACM; 2012. p. 638\u201349.","DOI":"10.1145\/2382196.2382263"},{"key":"2563_CR4","doi-asserted-by":"crossref","unstructured":"Zhang J, Cormode G, Procopiuc CM, Srivastava D, Xiao X. PrivBayes: Private Data Release via Bayesian Networks. In: Proc. 2014 ACM SIGMOD Int. Conf. Manage. Data. SIGMOD \u201914. New York: ACM; 2014. p. 1423\u201334.","DOI":"10.1145\/2588555.2588573"},{"key":"2563_CR5","doi-asserted-by":"crossref","unstructured":"Acs G, Melis L, Castelluccia C, De Cristofaro E. Differentially Private Mixture of Generative Neural Networks. IEEE Trans Knowl Data Eng. 2019;31(6):1109-21.","DOI":"10.1109\/TKDE.2018.2855136"},{"key":"2563_CR6","doi-asserted-by":"crossref","unstructured":"Abay NC, Zhou Y, Kantarcioglu M, Thuraisingham B, Sweeney L. Privacy Preserving Synthetic Data Release Using Deep Learning. In: Joint Eur. Conf. Mach. Learn. and Knowl. Discovery in Databases. Cham: Springer; 2018. p. 510\u2013526.","DOI":"10.1007\/978-3-030-10925-7_31"},{"key":"2563_CR7","unstructured":"Jordon J, Yoon J, van der Schaar M. PATE-GAN: Generating Synthetic Data with Differential Privacy Guarantees. In: 2019 Int. Conf. Learn. Representations. Appleton: ICLR; 2019."},{"key":"2563_CR8","unstructured":"McKenna R, Sheldon D, Miklau G. Graphical-model Based Estimation and Inference for Differential Privacy. In: Chaudhuri K, Salakhutdinov R, editors. Proc. 36th Int. Conf. Mach. Learn. PMLR; 2019. p. 4435\u201344."},{"key":"2563_CR9","doi-asserted-by":"publisher","unstructured":"Beaulieu-Jones BK, Wu ZS, Williams C, Lee R, Bhavnani SP, Byrd JB, et al. Privacy-preserving Generative Deep Neural Networks Support Clinical Data Sharing. Circ Cardiovasc Qual Outcomes. 2019;12(7). https:\/\/doi.org\/10.1161\/CIRCOUTCOMES.118.005122.","DOI":"10.1161\/CIRCOUTCOMES.118.005122"},{"key":"2563_CR10","unstructured":"Harder F, Adamczewski K, Park M. DP-MERF: Differentially Private Mean Embeddings with Random Features for Practical Privacy-preserving Data Generation. In: Int. Conf. Artif. Intell. and Statist. PMLR; 2021. pp. 1819\u201327."},{"key":"2563_CR11","doi-asserted-by":"publisher","unstructured":"J\u00e4lk\u00f6 J, Lagerspetz E, Haukka J, Tarkoma S, Honkela A, Kaski S. Privacy-preserving Data Sharing via Probabilistic Modeling. Patterns. 2021;2(7). https:\/\/doi.org\/10.1016\/j.patter.2021.100271.","DOI":"10.1016\/j.patter.2021.100271"},{"key":"2563_CR12","unstructured":"R\u00e4is\u00e4 O, J\u00e4lk\u00f6 J, Kaski S, Honkela A. Noise-Aware Statistical Inference with Differentially Private Synthetic Data. In: Proc. 26th Int. Conf. Artif. Intell. and Statist. PMLR; 2023. p. 3620\u201343."},{"key":"2563_CR13","unstructured":"Wilde H, Jewson J, Vollmer S, Holmes C. Foundations of Bayesian Learning from Synthetic Data. In: Banerjee A, Fukumizu K, editors. Proc. 24th Int. Conf. Artif. Intell. and Statist. PMLR; 2021. p. 541\u20139."},{"key":"2563_CR14","unstructured":"McMahan B, Moore E, Ramage D, Hampson S, y\u00a0Arcas BA. Communication-efficient Learning of Deep Networks from Decentralized Data. In: Proc. 20th Int. Conf. Artif. Intell. and Statist. PMLR; 2017. pp. 1273\u201382."},{"issue":"1\u20132","key":"2563_CR15","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1561\/2200000083","volume":"14","author":"P Kairouz","year":"2021","unstructured":"Kairouz P, McMahan HB, Avent B, Bellet A, Bennis M, Bhagoji AN, et al. Advances and Open Problems in Federated Learning. Found Trends Mach Learn. 2021;14(1\u20132):1\u2013210.","journal-title":"Found. Trends Mach. Learn."},{"key":"2563_CR16","doi-asserted-by":"crossref","unstructured":"Hsu J, Huang Z, Roth A, Wu ZS. Jointly Private Convex Programming. In: Proc. 27th Annu. ACM-SIAM Symp. Discrete Algorithms. Philadelphia: Society for Industrial and Applied Mathematics; 2016. p. 580\u201399.","DOI":"10.1137\/1.9781611974331.ch43"},{"key":"2563_CR17","doi-asserted-by":"publisher","unstructured":"Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. UK Biobank: An Open Access Resource for Identifying the Causes of a Wide Range of Complex Diseases of Middle and Old Age. PLoS Med. 2015;12(3). https:\/\/doi.org\/10.1371\/journal.pmed.1001779.","DOI":"10.1371\/journal.pmed.1001779"},{"key":"2563_CR18","doi-asserted-by":"crossref","unstructured":"Kifer D, Machanavajjhala A. No free lunch in data privacy. In: Proc. 2011 ACM SIGMOD Int. Conf. Manage. Data. New York: ACM; 2011. p. 193\u2013204.","DOI":"10.1145\/1989323.1989345"},{"key":"2563_CR19","doi-asserted-by":"crossref","unstructured":"Narayanan A, Shmatikov V. Robust De-anonymization of Large Sparse Datasets. In: 2008 IEEE Symposium on Security and Privacy. IEEE; 2008. p. 111\u201325.","DOI":"10.1109\/SP.2008.33"},{"key":"2563_CR20","unstructured":"Hu S, Wu S, Smith V. Private Multi-Task Learning: Formulation and Applications to Federated Learning. Trans Mach Learn Res. 2023."},{"key":"2563_CR21","doi-asserted-by":"publisher","unstructured":"Niedzwiedz CL, O\u2019Donnell CA, Jani BD, Demou E, Ho FK, Celis-Morales C, et al. Ethnic and Socioeconomic Differences in SARS-CoV-2 Infection: Prospective Cohort Study Using UK Biobank. BMC Med. 2020;18. https:\/\/doi.org\/10.1186\/s12916-020-01640-8.","DOI":"10.1186\/s12916-020-01640-8"},{"key":"2563_CR22","doi-asserted-by":"crossref","unstructured":"Seabold S, Perktold J. statsmodels: Econometric and statistical modeling with Python. In: Proc. 9th Python in Sci. Conf. Austin: SciPy.org; 2010. p. 92\u20136.","DOI":"10.25080\/Majora-92bf1922-011"},{"key":"2563_CR23","unstructured":"J\u00e4lk\u00f6 J, Dikmen O, Honkela A. Differentially Private Variational Inference for Non-conjugate Models. In: Proc. 33rd Conf. Uncertainty in Artif. Intell. Red Hook: Curran Associates, Inc; 2017."},{"key":"2563_CR24","unstructured":"J\u00e4lk\u00f6 J, Prediger L, Honkela A, Kaski S. DPVIm: Differentially Private Variational Inference Improved. Trans Mach Learn Res. 2023."},{"issue":"2","key":"2563_CR25","first-page":"407","volume":"2022","author":"L Prediger","year":"2022","unstructured":"Prediger L, Loppi N, Kaski S, Honkela A. d3p - A Python Package for Differentially-Private Probabilistic Programming. Proc Priv Enh Technol. 2022;2022(2):407\u201325.","journal-title":"Proc. Priv. Enh. Technol."},{"key":"2563_CR26","doi-asserted-by":"publisher","unstructured":"Song S, Chaudhuri K, Sarwate AD. Stochastic gradient descent with differentially private updates. In: IEEE Global Conf. Signal and Inf. Process, GlobalSIP. IEEE; 2013. p. 245\u20138. https:\/\/doi.org\/10.1109\/GlobalSIP.2013.6736861.","DOI":"10.1109\/GlobalSIP.2013.6736861"},{"key":"2563_CR27","doi-asserted-by":"crossref","unstructured":"Bassily R, Smith A, Thakurta A. Private Empirical Risk Minimization: Efficient Algorithms and Tight Error Bounds. In: 2014 IEEE Annu. Symp. Foundations Comput. Sci. IEEE; 2014. p. 464-73.","DOI":"10.1109\/FOCS.2014.56"},{"key":"2563_CR28","doi-asserted-by":"publisher","unstructured":"Abadi M, Chu A, Goodfellow I, McMahan HB, Mironov I, Talwar K, et al. Deep Learning with Differential Privacy. In: Proc. 2016 ACM SIGSAC Conf. Comput. Commun. Secur. CCS \u201916. New York: ACM; 2016. p. 308\u201318. https:\/\/doi.org\/10.1145\/2976749.2978318.","DOI":"10.1145\/2976749.2978318"},{"key":"2563_CR29","unstructured":"Koskela A, J\u00e4lk\u00f6 J, Honkela A. Computing Tight Differential Privacy Guarantees Using FFT. In: Chiappa S, Calandra R, editors. 23rd Int. Conf. Artif. Intell. and Statist. PMLR; 2020. p. 2560\u20139."},{"key":"2563_CR30","unstructured":"Koskela A, J\u00e4lk\u00f6 J, Prediger L, Honkela A. Tight Differential Privacy for Discrete-Valued Mechanisms and for the Subsampled Gaussian Mechanism Using FFT. In: Banerjee A, Fukumizu K, editors. 24th Int. Conf. Artif. Intell. and Statist. PMLR; 2021. p. 3358\u201366."},{"key":"2563_CR31","unstructured":"Breugel BV, Qian Z, Schaar MVD. Synthetic Data, Real Errors: How (Not) to Publish and Use Synthetic Data. In: Proc. 40th Int. Conf. Mach. Learn. PMLR; 2023. p. 34793\u2013808."},{"key":"2563_CR32","doi-asserted-by":"crossref","unstructured":"Rubin DB. Multiple Imputation for Nonresponse in Surveys. Hoboken: Wiley; 1987.","DOI":"10.1002\/9780470316696"},{"issue":"480","key":"2563_CR33","doi-asserted-by":"publisher","first-page":"1462","DOI":"10.1198\/016214507000000932","volume":"102","author":"JP Reiter","year":"2007","unstructured":"Reiter JP, Raghunathan TE. The Multiple Adaptations of Multiple Imputation. J Am Stat Assoc. 2007;102(480):1462\u201371.","journal-title":"J. Am. Stat. Assoc."},{"issue":"1","key":"2563_CR34","first-page":"1","volume":"19","author":"TE Raghunathan","year":"2003","unstructured":"Raghunathan TE, Reiter JP, Rubin DB. Multiple imputation for statistical disclosure limitation. J Off Stat. 2003;19(1):1.","journal-title":"J. Off. Stat."},{"issue":"1","key":"2563_CR35","first-page":"28","volume":"34","author":"BL Welch","year":"1947","unstructured":"Welch BL. The Generalization of \u2018Student\u2019s\u2019 Problem when Several Different Population Variances are Involved. Biometrika. 1947;34(1):28\u201335.","journal-title":"Biometrika."},{"issue":"3","key":"2563_CR36","doi-asserted-by":"publisher","first-page":"523","DOI":"10.1037\/h0078850","volume":"47","author":"DW Zimmerman","year":"1993","unstructured":"Zimmerman DW, Zumbo BD. Rank Transformations and the Power of the Student t Test and Welch t\u2019 Test for Non-normal Populations with Unequal Variances. Can J Exp Psychol. 1993;47(3):523\u201339.","journal-title":"Can. J. Exp. Psychol."},{"key":"2563_CR37","doi-asserted-by":"crossref","unstructured":"Yao AC. Protocols for Secure Computations. In: Proc. 23rd Annu. Symp. Found. Comput. Sci. IEEE; 1982. p. 160\u20134.","DOI":"10.1109\/SFCS.1982.38"},{"key":"2563_CR38","doi-asserted-by":"crossref","unstructured":"Lindell Y. Secure Multiparty Computation. Commun ACM. 2021;64(1):86-96","DOI":"10.1145\/3387108"},{"key":"2563_CR39","unstructured":"McMahan B, Ramage D, Talwar K, Zhang L. Learning Differentially Private Recurrent Language Models. In: 2018 Int. Conf. on Learn. Representations. Appleton: ICLR; 2018."},{"issue":"15","key":"2563_CR40","doi-asserted-by":"publisher","first-page":"3454","DOI":"10.1109\/TIFS.2020.2988575","volume":"1","author":"K Wei","year":"2020","unstructured":"Wei K, Li J, Ding M, Ma C, Yang HH, Farokhi F, et al. Federated Learning with Differential Privacy: Algorithms and Performance Analysis. IEEE Trans Info For Sec. 2020;1(15):3454\u201369.","journal-title":"IEEE Trans. Info. For. Sec."},{"key":"2563_CR41","unstructured":"Bietti A, Wei CY, Dudik M, Langford J, Wu S. Personalization Improves Privacy-Accuracy Tradeoffs in Federated Learning. In: Chaudhuri K, Jegelka S, Song L, Szepesvari C, Niu G, Sabato S, editors. Proc. 39th Int. Conf. Mach. Learn. PMLR; 2022. p. 1945\u201362."},{"key":"2563_CR42","unstructured":"Mirzasoleiman B, Bilmes J, Leskovec J. Coresets for Data-efficient Training of Machine Learning Models. In: Proc. 37th Int. Conf. Mach. Learn. PMLR; 2020. pp. 6950\u201360."},{"key":"2563_CR43","doi-asserted-by":"crossref","unstructured":"Xiong Y, Wang R, Cheng M, Yu F, Hsieh CJ. FedDM: Iterative Distribution Matching for Communication-Efficient Federated Learning. In: Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recog. IEEE; 2023. pp. 16323\u201332.","DOI":"10.1109\/CVPR52729.2023.01566"}],"container-title":["BMC Medical Informatics and Decision Making"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s12911-024-02563-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s12911-024-02563-7\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s12911-024-02563-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,6,14]],"date-time":"2024-06-14T16:03:10Z","timestamp":1718380990000},"score":1,"resource":{"primary":{"URL":"https:\/\/bmcmedinformdecismak.biomedcentral.com\/articles\/10.1186\/s12911-024-02563-7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,6,14]]},"references-count":43,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2024,12]]}},"alternative-id":["2563"],"URL":"https:\/\/doi.org\/10.1186\/s12911-024-02563-7","relation":{},"ISSN":["1472-6947"],"issn-type":[{"type":"electronic","value":"1472-6947"}],"subject":[],"published":{"date-parts":[[2024,6,14]]},"assertion":[{"value":"30 September 2023","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"3 June 2024","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 June 2024","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"UK Biobank has approval from the North West Multi-centre Research Ethics Committee (MREC) (21\/NW\/0157). All participants provided written informed consent before enrolment in the study, which was conducted in accordance with the Declaration of Helsinki.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethics approval and consent to participate"}},{"value":"Not applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}},{"value":"The authors declare no competing interests.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"167"}}