xpertsystems/hc-end-005-sample
收藏资源简介:
HC-END-005代谢综合征合成数据集(样本)是一个基于高斯copula(使用MESA/NHANES校准的相关性矩阵)构建的统计上复杂的代谢综合征患者合成队列。该数据集确保代谢综合征的五个核心组分——腰围、甘油三酯、高密度脂蛋白、空腹血糖和血压——是联合相关的,而非独立采样。数据集涵盖代谢综合征诊断(NCEP-ATP III和IDF标准)、连续严重程度评分(Gurka/Kaplan Z-score)、4类分期、完整的血脂/血糖/血压面板、心血管疾病风险(PCE/SCORE2/Framingham评分)、脂肪因子与炎症指标、肾脏/肝脏标志物、生活方式、干预组(基于糖尿病预防计划DPP风格)以及结果。此存储库包含一个500行、单种子的样本。完整商业产品可扩展至30,000多名患者,提供15年随访数据,并以CSV、Parquet和JSON格式交付。
A statistically sophisticated synthetic cohort of metabolic-syndrome patients built on a Gaussian copula (MESA/NHANES-calibrated correlation matrix) so the five MetS components — waist, triglycerides, HDL, fasting glucose, and blood pressure — are jointly correlated rather than independently sampled. Covers MetS diagnosis (NCEP-ATP III & IDF), continuous severity (Gurka/Kaplan Z-score), 4-cluster staging, full lipid/glycemic/BP panels, CVD risk (PCE/SCORE2/Framingham), adipokines & inflammation, renal/hepatic markers, lifestyle, intervention arms (DPP-style), and outcomes. This repository contains a 500-row, single-seed sample. The full commercial product scales to 30,000+ patients with 15-year follow-up and CSV / Parquet / JSON delivery.




