题 目:Robust Variable Selection of Joint Frailty Model for Panel Count Data
主 讲 人:赵晓兵教授
主 持 人:罗季副教授
时 间:2016年10月11号(周二)14:40-15:40
地 点:yh86银河国际楼(6号学院楼)415教室
主讲人简介:
赵晓兵教授,获香港理工大学哲学博士学位,华东师范大学博士后。现为yh86银河国际教授,浙江省高校中青年学科带头人。主要研究方向为生存分析、复发事件数据分析、高维数据分析,以及保险精算中相关问题。在统计学、保险精算等领域主流期刊Statistica Sinica, Journal of Multivariate Analysis, Statistics in Medicine, Insurance:Mathematics and Economics等期刊发表论文30余篇;主持国家自然科学基金2项、浙江省自然科学基金2项;多次到澳大利亚麦考瑞大学、美国西北大学进行学术访问。
摘 要:
Panel count data arise when the subjects are examined or observed at discrete time points of the event history. Analysis of panel count data has recently received a great deal of attention in the literature. Low-dimensional covariates have been commonly assumed in the existing literature. In practice, however, the panel count data may often involve high-dimensional covariates. The literature on panel count data with high-dimensional covariates has so far been quite limited. In this paper, a robust variable selection approach based on the quantile regression function in a joint frailty model is proposed to analyze panel count data with high-dimensional covariates. A three-step estimation procedure is developed to select variables and estimate the regression coefficients. The consistency and oracle property are established under some suitable conditions. Simulations are carried out to confirm and assess the performance of the proposed method and an example is demonstrated on a dataset from a bladder tumor study.
欢迎各位同学和老师踊跃参加!