时间：2017年06月23日周五，下午1:00—2:00
地点：统计楼105
报告人：印第安那大学生物统计系张殷教授
题目：Two-stage semiparametric analysis of skeletal growth around pubertal growth spurt with interval-censored observations
摘要：
Human individuals acquire their adult body shapes through vigorous physical growth in the first two decades of life. Many of the somatic characteristics that define our physical appearance in adulthood take shape around the time of pubertal growth spurt (PGS). An analytical challenge to quantify growth rates before and after PGS is the lack of direct observation of the anchoring PGS event. We propose a two-stage semiparametric analysis to assess the rates of skeletal changes around the PGS with interval-censored observation on the PGS. The first stage is the nonparametric maximum likelihood estimation for the distribution of PGS timing. In the second stage, a least-squares based method is used to estimate the model parameters, including the pre and post-PGS growth rates with latent time of PGS. We show that under mild regularity conditions, the estimators are consistent and asymptotically normal. Statistical inference ensues from the large sample theory. We conduct a simulation study to evaluate the operating characteristics of the proposed method. Analysis of growth data from an observational cohort shows that in comparison to girls, boys tend to have a more sustained skeletal growth after PGS, as evidenced by the greater post-PGS growth rates in the upper body. The findings suggest that strong and sustained post-PGS skeletal growth contributes to the sexual dimorphism in human body.
报告人简介：
ZhangYing
Professor & Director of Education
Department of Biostatistics
IU Fairbanks School of Public Health and
IU School of Medicien
Academic Appointment
1988-1991, Lecturer, Department of Mathematics, Fudan University
1998-2004, Assistant Professor, Department of Statistics, University of Central Florida
2004-2010, Associate Professor, Department of Biostatistics, University of Iowa
2010-2014, Professor, Department of Biostatistics, University of Iowa
2014-present, Professor, Department of Biostatistics, Indiana University
Research
Dr. Zhang has a broad interest in statistical/biostatistical methodology research including non-/semi-parametric inference, panel count data analysis, survival data analysis, joint modeling of survival and longitudinal data, clinical trials, statistical computing and data mining. Many of his research results appear in top-tier statistical journals such as Annals of Statistics, Journal of the American Statistical Association, Biometrika and Bometrics. In addition to methodological research, Dr. Zhang is also actively engaged in NIH or CDC funded collaborative research with scientists in the areas of nursing, social work, sports medicine, epidemiology, infectious diseases and psychiatrics. Currently, he is heavily involved in NIH/NINDS funded project PREDICT-HD as a site PI to study the early progression of Huntington disease aiming to discover progression biomarkers for this neurodegenerative disease.