报告人：孙六全 (中国科学院数学与系统科学研究院)

邀请人：潘灯

报告时间：2022年4月6日 （星期三）10: 00-12:00  

报告地点：腾讯会议：693 853 701

报告题目：Variable Screening for Several Varying Coefficient Models with Ultrahigh-Dimensional Survival Data

报告摘要：In this talk, we develop variable screening methods for several varying coefficient hazards models, including varying coefficient additive hazards model, varying coefficient Cox model, and the varying coefficient hazards models of single-index form. The proposed methods enjoy the sure screening property, and some iterative procedures are developed. Simulation studies are conducted to evaluate the finite sample performances of the proposed method. Applications to real data examples are also provided. 

报告人简介：孙六全，中国科学院数学与系统科学研究院研究员，博士生导师，中国科学院数学与系统科学研究院统计中心副主任。主要研究方向为生存分析、生物统计、纵向数据和复发事件数据以及复杂删失数据的统计分析。在JASA，Biometrik，JMLR及SMMR等国内外核心期刊发表学术论文170余篇。