用私有微调大模型生成时序表格数据，更好保留事件先后关系。
Privately Fine-Tuned LLMs Preserve Temporal Dynamics in Tabular Data
- 以整张表为单位生成，利用大模型自回归能力建模时间序列。
- 相比主流方法，轨迹分布距离降低60%以上，状态转移错误减少近50%。
- 适合医疗记录等含时间顺序的敏感数据合成，兼顾隐私与结构保真。
关于差分隐私合成表格数据的研究大多聚焦于独立同分布的行，每行对应一个独立个体。这种视角忽略了纵向数据（如电子健康记录）中的时间复杂性——用户会贡献一系列连续事件组成的（子）表。尽管从业者可能将用户历史展平为高维向量并用于标准边际机制，我们证明该策略不足：展平虽能保持有效的边际分布，却无法保留时间连贯性。为此，我们提出PATH框架，将整张表作为合成单元，利用私有微调的大语言模型的自回归能力。大量实验表明，PATH有效捕捉了传统方法遗漏的长程依赖。实证结果显示，与领先边际机制相比，该方法使真实轨迹的分布距离降低超过60%，状态转移错误减少近50%，同时保持相当的边际保真度。
原文摘要 · Abstract (English)
Research on differentially private synthetic tabular data has largely focused on independent and identically distributed rows where each record corresponds to a unique individual. This perspective neglects the temporal complexity in longitudinal datasets, such as electronic health records, where a user contributes an entire (sub) table of sequential events. While practitioners might attempt to model such data by flattening user histories into high-dimensional vectors for use with standard marginal-based mechanisms, we demonstrate that this strategy is insufficient. Flattening fails to preserve temporal coherence even when it maintains valid marginal distributions. We introduce PATH, a novel generative framework that treats the full table as the unit of synthesis and leverages the autoregressive capabilities of privately fine-tuned large language models. Extensive evaluations show that PATH effectively captures long-range dependencies that traditional methods miss. Empirically, our method reduces the distributional distance to real trajectories by over 60% and reduces state transition errors by nearly 50% compared to leading marginal mechanisms while achieving similar marginal fidelity.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。