Robust Estimation in Ignorable Missing data
主 题: Robust Estimation in Ignorable Missing data
报告人: 殷云剑博士 (北京大学数学学院)
时 间: 2015-10-22 12:00-13:15
地 点: 理科一号楼1560
In the presence of a randomly missing response, two different approaches can be employed to estimate the mean of the response variable, one is to reweight the complete case subsample by the inverse of nonmissing probability, another is to fit a regression model for the conditional mean of the response given covariates. Both approaches have merits but fail to be consistent when their underlying model is misspecified. In the talk we introduce some robust estimators in the literature, the robust means the estimator remains consistent when only one of the models is correctly specified. 数学学院研究生学术午餐会简介:学院领导的大力支持,由院研究生会负责组织的系列学术交流活动。午餐会每次邀请一位同学作为主讲人,面向全院各专业背景的研究生介绍自己科研方向的基本问题、概念和方法,并汇报近期的研究成果和进展,是研究生展示自我、促进交流的学术平台,已经举办了四期活动。研究生会将于2015年10月8日周四举办第五期学术午餐会活动,欢迎感兴趣的老师和同学积极报名参加。 请有意参加的老师发送邮件至pkusmsgu@163.com,我们将回复邮件和您进行确认。