主 题：Missing Covariates in Quantile Regression.
主讲人：Ying Wei副教授
主持人：林华珍教授
时 间：2014年10月24日16：30-17：10
地 点：通博楼B座212学术会议室
主办单位：统计学院 统计研究中心 科研处
主讲人简介：
Ying Wei is a tenured Associate Professor of Biostatistics of Columbia University. She received her PhD in 2004 from the University of Illinois at Urbana-Champaign. Her research interests include quantile regression, longitudinal data analysis, measurement errors, and growth charts. She received the 2011 Noether Young Scholar Award from the American Statistical Association for her significant contributions to the advancement of nonparametric statistics. Currently she is the director of the Consulting Center at the Biostatistics Department at Columbia, and serves on the editorial board of JASA. She has published discussion papers in Annals of Statistics and in The American Journal of Epidemiology.
内容提要：
Regression quantiles can be underpowered or biased when there are missing values in some covariates. Depending on the missing data mechanism, we develop several approaches to handle missing covariates to correct potential bias and achieve a better efficiency. The finite sample performances of our estimators are investigated through simulation studies. Finally, to illustrate the utility of the proposed methods, we apply our methodology to a national nutritional survey data on dietary intakes.
