Department of Statistics, Renmin University of China, Beijing, China.
School of Statistics, Lanzhou University of Finance and Economics, Gansu, China.
Stat Med. 2019 Oct 15;38(23):4670-4685. doi: 10.1002/sim.8326. Epub 2019 Jul 29.
The proportional subdistribution hazard regression model has been widely used by clinical researchers for analyzing competing risks data. It is well known that quantile regression provides a more comprehensive alternative to model how covariates influence not only the location but also the entire conditional distribution. In this paper, we develop variable selection procedures based on penalized estimating equations for competing risks quantile regression. Asymptotic properties of the proposed estimators including consistency and oracle properties are established. Monte Carlo simulation studies are conducted, confirming that the proposed methods are efficient. A bone marrow transplant data set is analyzed to demonstrate our methodologies.
比例亚分布风险回归模型已被临床研究人员广泛用于分析竞争风险数据。众所周知，分位数回归为模型如何影响协变量不仅提供了一个更全面的替代位置，而且还提供了对整个条件分布的影响。在本文中，我们针对竞争风险分位数回归开发了基于惩罚估计方程的变量选择程序。建立了所提出估计量的渐近性质，包括一致性和 oracle 性质。通过蒙特卡罗模拟研究，证实了所提出的方法是有效的。分析了一个骨髓移植数据集，以证明我们的方法。
