合成数据能否复现流行病学真实发现？基于对抗随机森林的复现研究
Can synthetic data reproduce real-world findings in epidemiology? A replication study using adversarial random forests
研究用对抗随机森林（ARF）合成表格型流行病学数据，复现了六篇流行病学论文的统计分析，涵盖血压、人体测量、心肌梗死、加速度计、孤独感和糖尿病，数据来自德国国家队列（NAKO）、不来梅 STEMI 登记 U45 研究和 Guelph 家庭健康研究。ARF 合成数据在描述性与推断性分析中均与原始结果一致，即便样本量-维度比偏低时也如此，且在效用、隐私保护、泛化性和运行时间上优于常用表格数据合成器。
Authors:Jan Kapar, Kathrin Günther, Lori Ann Vallis, Klaus Berger, Nadine Binder, Hermann Brenner, Stefanie Castell, Beate Fischer, Volker Harth, Bernd Holleczek, Timm Intemann, Till Ittermann, André Karch, Thomas Keil, Lilian Krist, Berit Lange, Michael F. Leitzmann, Katharina Nimptsch, Nadia Obi, Iris Pigeot, Tobias Pischon, Tamara Schikowski, Börge Schmidt, Carsten Oliver Schmidt, Anja M. Sedlmair, Justine Tanoey, Harm Wienbergen, Andreas Wienke, Claudia Wigmann, Marvin N. Wright
Abstract:Synthetic data hold substantial potential to address practical challenges in epidemiology due to restricted data access and privacy concerns. However, many current methods suffer from limited quality, high computational demands, and complexity for non-experts. Furthermore, common evaluation strategies for synthetic data often fail to directly reflect statistical utility and measure privacy risks sufficiently. Against this background, a critical underexplored question is whether synthetic data can reliably reproduce key findings from epidemiological research while preserving privacy. We propose adversarial random forests (ARF) as an efficient and convenient method for synthesizing tabular epidemiological data. To evaluate its performance, we replicated statistical analyses from six epidemiological publications covering blood pressure, anthropometry, myocardial infarction, accelerometry, loneliness, and diabetes, from the German National Cohort (NAKO Gesundheitsstudie), the Bremen STEMI Registry U45 Study, and the Guelph Family Health Study. We further assessed how dataset dimensionality and variable complexity affect the quality of synthetic data, and contextualized ARF's performance by comparison with commonly used tabular data synthesizers in terms of utility, privacy, generalization, and runtime. Across all replicated studies, results on ARF-generated synthetic data consistently aligned with original findings. Even for datasets with relatively low sample size-to-dimensionality ratios, replication outcomes closely matched the original results across descriptive and inferential analyses. Reduced dimensionality and variable complexity further enhanced synthesis quality. ARF demonstrated favourable performance regarding utility, privacy preservation, and generalization relative to other synthesizers and superior computational efficiency.
| Subjects: | Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Applications (stat.AP); Machine Learning (stat.ML) |
| Cite as: | arXiv:2508.14936 [q-bio.QM] |
| (or arXiv:2508.14936v4 [q-bio.QM] for this version) | |
| https://doi.org/10.48550/arXiv.2508.14936 arXiv-issued DOI via DataCite |
|
| Journal reference: | International Journal of Epidemiology, 55(5), dyag164 (2026) |
| Related DOI: | https://doi.org/10.1093/ije/dyag164
DOI(s) linking to related resources |
Submission history
From: Jan Kapar [view email]
[v1]
Tue, 19 Aug 2025 22:51:40 UTC (113 KB)
[v2]
Mon, 23 Mar 2026 11:45:17 UTC (175 KB)
[v3]
Tue, 5 May 2026 06:01:43 UTC (160 KB)
[v4]
Wed, 21 Oct 2026 13:44:20 UTC (161 KB)
来源：arXiv cs.AI · arxiv.org