应数学系概率统计与运筹控制研究所邀请，美国印第安纳大学南本德分校关忠教授将于近日访问数学系并作报告。

报告题目：Semiparametric Models for Data with Nonresponses

报告时间：7月15日下午13:00-14:30,   14:30-16:00

                    7月19日上午8:30-10:00，  10:00-11:30

报告地点：格物楼503

报告摘要：I will start  this series of talks by introducing the problem in survey sampling with nonresponses. The nonresponses or missing values are classified as (i) Missing completely at random, i.e., data missing mechanism is independent of any  observable or unobservable quantities. (ii) Missing at random,i.e., data missing mechanism depends on only the observed variables.   (iii) Non-ignorable missing data, i.e., data missing mechanism depends on their own values. Missing at random is also called ignorable missing which means that we can ignore those subjects with missing at random values. However, the non-ignorable missing is the most difficult problem in the missing data literature. I will report our recent work on non-ignorable missing in a very general setting.In the presence of nonignorable nonresponses in a survey sampling problem with or without callbacks, a logistic regression using the missing response variable as one of the covariates is  adopted to express the response probability. The model is proved identifiable. Semiparametric maximum likelihood estimators of the parameters in the response probability are proposed and studied. As a result, an efficient estimator of the mean of the response variable is constructed using the estimated  response probability. Moreover, if a regression model for conditional mean of the response variable given some covariate is available, then also proposed is the method to obtain an even more efficient estimate of the mean of the response variable  by fitting the regression model using an adjusted least squares method based on the estimated underlying distributions of the observed values. Simulation results show the efficiency of the proposed methods compared with some existing competitors. An application to a Korean household survey of employment describes the usage of the method.

 关忠教授简介：关忠，美国印第安娜大学南本德分校 (University of Indianna at South Bend )教授。1978-1982年哈尔滨科技大学学习，1982-1985年002全讯白菜网硕士毕业，1985-1997年002全讯白菜网数学系统计与运筹教研室任讲师、副教授，1989-1990年曾在荷兰名校莱登大学进修一年；1997-2001年毕业于美国俄亥俄州托莱多大学，获统计学博士学位。2001-2004年在美国耶鲁大学医学院生物统计博士后，2004-至今在美国印第安娜大学南本德分校分别任助理教授、副教授并获终身教授职位。主要研究方向为生物统计、基因微阵列数据分析、非参数和半参数方法及其应用、经验似然方法及其应用、变点估计等。担任多个国际知名统计学SCI刊物审稿人。在Biometrika, Biometrics, Bioinformatics，Statistic Sinica, Canadian Journal of Statistics，Journal of Nonparametrics Statistics等国际顶尖统计学期刊上发表近20篇学术论文，在生物统计等领域有非常出色的工作；熟悉国内外统计学的教学和科研发展方向。

关忠博士留美前曾是002全讯白菜网数学系副教授，统计与运筹教研室副主任，曾担任002全讯白菜网学报英文版编委，数学系教学指导委员会委员等职。

欢迎数学系及其它相关专业师生参加本次学术报告!