统计与数据科学学院
-------------------------------------------------
学术讲座
讲座题目：Joint generalized estimating equation and its implementation into R
讲座人：潘建新教授，英国曼彻斯特大学数学学院
时间：2018年6月27日（周三）10:30
地点：统计与数据科学学院426教室
摘要: The approach of generalized estimating equations (GEE) is widely used in the analysis of clustered and longitudinal data. This technique requires a specification of working structure for the within-subject covariance matrix and can obtain consistent estimates of the mean parameters even if the working covariance structure is misspecified. However, this may result in a great loss of efficiency of the mean parameter estimators. In some circumstance, for example, when missing data occur, the misspecification can lead to biased estimators of the mean parameters. It is demonstrated that Cholesky-type decomposition-based methods are effective for modelling of the covariance structure and can improve statistical inference for the mean model. In this research, we first consider modified Cholesky decomposition (MCD) based methods for modelling the mean-covariance structures in the framework of generalized estimating equations (GEE), and then study the inverse probability weighted generalized estimating equations (WGEE) for handling missing data. We also introduce our newly developed R package gee4 which is able to handle continuous longitudinal data that follows a non-Normal distribution using the GEE-MCD and WGEE-MCD methods. Simulated and real datasets are used to illustrate the use of functions in this package.
(Joint work with Dr Yi Pan)
讲座人简历：潘建新教授，英国曼彻斯特大学数学学院终身教授，英国皇家统计学会会员, 国际统计学会当选会员和美国数理统计学会会员。 统计学杂志Biometrics和Biometrical Journal 编委（Associated Editor）。1996年在香港浸会大学获得统计学博士学位，之后到英国洛桑(Rothamsted)实验中心从事博士后研究。2002年10月加盟曼彻斯特大学数学学院，先后仼讲师(2002)、高级讲师(2004)、Reader(2005)。2006年被曼彻斯特大学聘为终身教授，并兼任曼彻斯特大学医学院研究员。曾担任曼大数学学院概率统计系系主任。致力于统计学领域内复杂数据模型的理论研究及其在医学及工业上的应用，取得了多项创新性研究成果。成果发表在包括Journal of the American Statistical Association和Biometrika在内的统计学主流期刊上。至今已发表学术论文100余篇，出版学术专著2部，其中1部于2002年由Springer出版社出版。已指导18名博士研究生并获得学位。
欢迎广大师生参加！
统计与数据科学学院
2018年6月13日