基于LASSO的生存预测建模与多重插补数据：结核病死亡率预测案例研究
基于多重插补数据的LASSO生存预测模型在结核病死亡率预测中的R代码与模拟数据。
基本信息
资源简介
基于LASSO的生存预测建模与多重插补数据：结核病死亡率预测案例研究
【一句话概览】
针对健康行政数据缺失预测变量的问题，本研究以结核病死亡率预测为例，探讨多重插补数据下Cox-LASSO生存预测建模。比较预测平均、性能平均和堆叠三种方法，以及minimum-lambda和1SE-lambda超参数选择。结果显示，堆叠法最稳健，时间依赖c-statistic达0.93，校准良好；1SE-lambda多致欠拟合。研究推荐堆叠法结合minimum-lambda，并共享可重复R代码（来源 figshare，CC BY 4.0 许可）。
【数据集背景】
本数据集源站标题为《LASSO-Based Survival Prediction Modeling with Multiply Imputed Data: A Case Study in Tuberculosis Mortality Prediction》。配套论文《LASSO-Based Survival Prediction Modeling with Multiply Imputed Data: A Case Study in Tuberculosis Mortality Prediction》发表于 The American Statistician（2025），DOI：10.1080/00031305.2025.2526545。项目资助方：This research received no external funding. MBH is supported by the University of British Columbia 4-Year Doctoral Fellowship (2020–2024) and Harry and Florence Dennison Fellowship in Medical Research (2021–2022, 2022–2023). MEK has received support from the Michael Smith Health Research BC Scholar Award, MS Canada Discovery grant (PG#: 1034713), and his research is backed by the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant (RGPIN-2018-05044) and Discovery Launch Supplement (DGECR-2018-00235). The researchers were independent of their sources of financial support and the sources of financial support had no role in this study.。源站首次发布：2025-08-04。
【数据内容】
源站原始描述：Using health administrative datasets for developing prediction models is always challenging due to missing values in key predictors. Multiple imputation has been recommended to deal with missing predictor values. However, predicting survival outcomes using regularized regression, for example, Cox-LASSO, faces limitations as these methods are incompatible with pooling model outputs from multiple imputed data using Rubin’s rule. In this study, we explored the performance of three statistical metho…
中文译文（机器翻译，仅供快速理解，请以原文为准）：使用卫生行政数据集开发预测模型始终具有挑战性，因为关键预测变量中存在缺失值。多重插补已被推荐用于处理预测变量缺失值。然而，使用正则化回归（例如 Cox-LASSO）预测生存结局面临局限性，因为这些方法无法使用 Rubin 规则合并来自多重插补数据的模型输出。在本研究中，我们探讨了在多重插补数据上使用 Cox-LASSO 开发预测模型时三种统计方法的性能：预测平均、性能平均和堆叠。我们考虑了两种超参数选择技术：给出最小交叉验证预测误差的 minimum-lambda
【数据结构与技术规格】
文件数 2 个，合计 968.68 KB。Supplementary material 1.pdf：application/pdf，803.36 KB。Supplementary material 2.pdf：application/pdf，165.32 KB。数据集 DOI：10.6084/m9.figshare.29447083.v2。
【主题与分类】
源站学科分类：Biophysics、Medicine、Genetics、Biological Sciences not elsewhere classified、Information Systems not elsewhere classified、Mathematical Sciences not elsewhere classified。主题标签：Internal validation、LASSO、Multiple imputation、Prediction、Survival outcome。
【适用方向】
结构化表格数据可直接用于统计建模、特征工程与队列分析。
【获取与许可】
源站页面：https://tandf.figshare.com/articles/dataset/LASSO-based_Survival_Prediction_Modelling_with_Multiply_Imputed_Data_A_Case_Study_in_Tuberculosis_Mortality_Prediction/29447083
使用许可：CC BY 4.0（https://creativecommons.org/licenses/by/4.0/）。
规范引用：LASSO-Based Survival Prediction Modeling with Multiply Imputed Data: A Case Study in Tuberculosis Mortality Prediction. Taylor & Francis. Hossain, Md. Belal; Sadatsafavi, Mohsen; Johnston, James C.; Wong, Hubert; Cook, Victoria J.; Karim, Mohammad Ehsanul
下载信息
注册下载
需要注册 Kaggle 账号并登录后下载，适合需要跟踪下载记录和使用 API 的用户。
暂未开放公开下载
无需注册即可直接获取公开样本或文档，适合快速预览和评估数据集质量。
免登录有偿下载
公开数据集受托下载与技术交付服务。
提供高速下载与技术交付服务（收技术服务费，非数据销售）
暂未开放当前数据集主要来源为 Kaggle 公开托管，完整影像包建议通过原始链接或 Kaggle API 获取。
使用方式
数据集获取（Figshare）
- 源站标题：LASSO-Based Survival Prediction Modeling with Multiply Imputed Data: A Case Study in Tuberculosis Mortality Prediction
- DOI：10.6084/m9.figshare.29447083.v2
- 发布日期：2025-08-04
- 文件数：2 个
- 文件总体积：968.68 KB
- 源站页面：https://tandf.figshare.com/articles/dataset/LASSO-based_Survival_Prediction_Modelling_with_Multiply_Imputed_Data_A_Case_Study_in_Tuberculosis_Mortality_Prediction/29447083
命令行下载
curl -L -o "Supplementary material 1.pdf" "https://ndownloader.figshare.com/files/55900414"
curl -L -o "Supplementary material 2.pdf" "https://ndownloader.figshare.com/files/55900417"
文件清单
Supplementary material 1.pdf（803.36 KB）Supplementary material 2.pdf（165.32 KB）
文件总体积 968.68 KB（源站 API 实测）。
数据加载示例（图像类）
from PIL import Image
import glob, os
files = (glob.glob(os.path.join(path, "**", "*.png"), recursive=True)
+ glob.glob(os.path.join(path, "**", "*.jpg"), recursive=True)
+ glob.glob(os.path.join(path, "**", "*.tif"), recursive=True))
print("图像文件数:", len(files))
img = Image.open(files[0]); print("尺寸/模式:", img.size, img.mode)
# torchvision Dataset 方式:
# from torchvision import datasets
# ds = datasets.ImageFolder(path) # 要求 子目录=类别
目录组织与标注格式以源站说明和下载后实际文件为准。
数据说明
Using health administrative datasets for developing prediction models is always challenging due to missing values in key predictors. Multiple imputation has been recommended to deal with missing predictor values. However, predicting survival outcomes using regularized regression, for example, Cox-LASSO, faces limitations as these methods are incompatible with pooling model outputs from multiple imputed data using Rubin’s rule. In this study, we explored the performance of three statistical methods in developing prediction models with Cox-LASSO on multiply imputed data: prediction average, performance average, and stacked. We considered two hyperparameter selection techniques: minimum-lambda that gives the minimum cross-validated prediction error and 1SE-lambda that selects more parsimonious models. We also conducted plasmode simulations with varying the events per parameter. The stacked approach provided the most robust predictions in our case study of predicting tuberculosis mortality and simulations, producing a time-dependent c-statistic of 0.93 and a well-calibrated calibration plot. The 1SE-lambda technique resulted in underfitting of the models in most scenarios, both in case study and simulation. Our findings advocate the stacked method with minimum-lambda as an effective technique for combining LASSO-based prediction outputs from multiply imputed data. We shared reproducible R codes for future researchers to facilitate the adoption of these methodologies in their research.
许可
CC BY 4.0
引用
学术使用请引用 DOI 10.6084/m9.figshare.29447083.v2（source: https://tandf.figshare.com/articles/dataset/LASSO-based_Survival_Prediction_Modelling_with_Multiply_Imputed_Data_A_Case_Study_in_Tuberculosis_Mortality_Prediction/29447083）。
数据缺失？
依托陕西、四川两大基地，我们与超过十家三甲医院建立直接合作关系，覆盖合规授权、采集治理、专业标注、数据交付的全流程，为AI医疗团队提供即拿即用的高质量临床数据。
⚡️ 需要数据支持或标注服务？立即联系我们获取专业方案。
