RawMed
收藏资源简介:
RawMed数据集是一个多表时间序列电子健康记录(EHR)数据集,由韩国科学技术院(KAIST)和FuriosaAI合作创建。该数据集通过文本表示和压缩技术,能够捕获复杂结构和时间动态,并且预处理步骤最少。RawMed旨在解决电子健康记录中敏感个人信息的隐私问题,通过生成合成EHR数据集来支持医疗研究和应用。该数据集在两个开源EHR数据集上进行了验证,证明了其在保真度和实用性方面的优势。
RawMed is a multi-table time-series electronic health record (EHR) dataset jointly created by the Korea Advanced Institute of Science and Technology (KAIST) and FuriosaAI. This dataset adopts text representation and compression technologies to capture complex structures and temporal dynamics, with minimal preprocessing steps. RawMed aims to resolve the privacy issues of sensitive personal information in electronic health records, and supports medical research and applications by generating synthetic EHR datasets. The dataset has been validated on two open-source EHR datasets, proving its advantages in terms of fidelity and practical usability.
RawMed数据集概述
基本信息
- 数据集名称:RawMed

- 1Generating Multi-Table Time Series EHR from Latent Space with Minimal Preprocessing韩国科学技术院(KAIST) · 2025年



