文：西南财经大学统计学院 发布时间：2016-05-09 浏览次数：1011
主 题：A Nonparametric Regression Model for Panel Count Data Analysis
主办单位：统计研究中心 统计学院 科研处
Dr. Yu is currently a Professor of Biostatisitcs at SJTU-Yale Joint Center for Biostatistics, Shanghai Jiaotong University. He earned his BS degree of Statistics from East China Normal University, MA degree of Statistics from Bowling Green State University, Phd degree of Biostatistics from University of Michigan. Previously, he has been an associate professor in the Department of Biostatistics, Indiana University School of Medicine. His methodology researches focus on survival data analysis, longitudinal data analysis, variable selection, nonparametric regression, variable selection. He also collaborate medical investigators on pediatric pulmonology, cancer, hypertension, health service research. He serves as an associate editor of Statistics in Medicine, Journal of System Science and Complexity and editorial boards of multiple clinical journals.
Panel count data are commonly encountered in event time studies where the exact event occurrence times are unobserved which are commonly encountered in oncology studies but conveniently ignored. To accommodate the potential non-linear covariate effect, we consider a nonparametric regression model for panel count data. Regression spline method is used to estimate the nonparametric covariate functions and the baseline mean function. The estimators are showed to be consistent and achieve the optimal rate for spline estimators. Numerical studies show that the estimators have good finite sample properties too. We apply the proposed method to analyze the non-linear effect of the lung function on the risks of childhood wheezing.