Statistics and Its Interface
Volume 14 (2021)
Group variable selection for recurrent event model with a diverging number of covariates
Pages: 431 – 447
For the high-dimensional data, the number of covariates can be large and diverge with the sample size. In this work, we propose an adaptive bi-level penalized method to solve the group variable selection problem for the recurrent event model with a diverging number of covariates. Comparing with the classical group variable selection methods, the adaptive bi-level penalized method can select the important group variables and individual variables simultaneously. For the case of diverging a number of covariates, we demonstrate that the proposed method has selection consistency and the penalized estimators have asymptotic normality. Simulation studies show that the proposed method performs well and the results are consistent with the theoretical properties. The proposed method is illustrated by analyzing a real life data set.
group variable selection, recurrent event model, adaptive bi-level penalty, diverging dimension, asymptotic normality, oracle property
2010 Mathematics Subject Classification
Primary 62J07. Secondary 62F12, 62N01, 62P10.
Shen’s research is partially supported by a Discovery Grant (RG/PIN04594-2016) from Natural Sciences and Engineering Research Council (NSERC) of Canada and a funding from Breast Cancer Society of Canada.
Lu’s research is partially supported by a Discovery Grant (RG/PIN06466-2018) from NSERC.
Received 12 April 2020
Accepted 29 January 2021
Published 8 July 2021