arXiv：cs.LG· Talal Alrawajfeh, Cristiana Diaconu, Ossi R\"ais\"a, Sebastian Rodriguez Beltran, Yuan He, John Bronskill, Richard E. Turner, Antti Honkela·· 4 小时前AI 评分43
PrivTab：一种可证明隐私保护的表格基础模型分类方法
Efficient Provably Private Classification with a Tabular Foundation Model
AI 导读
研究人员提出 PrivTab，一种将隐私机制嵌入架构的差分隐私分类表格基础模型，在模拟数据集上预训练，通过上下文学习把敏感行转化为可证明隐私的紧凑摘要。该模型在中高强度隐私下优于私有线性与神经网络基线，成员泄漏可忽略、预测校准良好，并将数据集拟合时间缩短 10000 倍，仅需一次前向传播。
正文
Abstract:Tabular data underpin prediction and decision-making in medicine, finance, government and science, but often contain sensitive individual-level information, creating a need for accurate prediction while preserving privacy. Traditional private learning provides formal privacy guarantees, but requires slow dataset-specific optimisation, suffers substantial utility loss under strong privacy, and is often difficult to apply correctly. Tabular foundation models adapt rapidly to new datasets, but existing models lack formal privacy guarantees, and are highly vulnerable to membership-inference attacks, limiting their use on sensitive data. Here we introduce PrivTab, an easy to use tabular foundation model for differentially private classification that embeds a privacy mechanism within its architecture. Pretrained on simulated datasets, PrivTab uses in-context learning to transform sensitive rows into compact, provably private summaries---effectively learning how to learn under privacy. PrivTab outperforms private linear and neural-network baselines under moderate-to-strong privacy, shows negligible membership leakage, maintains well-calibrated predictions under strong privacy, and reduces dataset fitting time by 10,000 times, requiring only a single forward pass. By combining formal privacy, speed, and easy of use, PrivTab brings recent advances in AI to applications where sensitive individual-level data have limited their adoption.
| Comments: | 74 pages, 18 figures; includes supplementary information |
| Subjects: | Machine Learning (cs.LG); Cryptography and Security (cs.CR) |
| Cite as: | arXiv:2610.10068 [cs.LG] |
| (or arXiv:2610.10068v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.10068 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Talal Alrawajfeh [view email]
[v1]
Wed, 7 Oct 2026 13:34:38 UTC (26,853 KB)
来源：arXiv：cs.LG · arxiv.org