有偏抽样下带终止时间和带信息观察时间的纵向数据分析  

Analysis of Longitudinal Data with Informative Observation and Terminal Event Times Under Biased Sampling

作　　者：苗瑞[1] 李怿[2] 

机构地区：[1]中国科学院数学与系统科学研究院,北京100190 [2]中国科学技术大学统计与金融系,合肥230026

出　　处：《应用数学学报》2013年第6期961-977,共17页Acta Mathematicae Applicatae Sinica

基　　金：国家自然科学基金(11231010;11171330;11021161)资助项目

摘　　要：纵向数据经常出现在生物医学等实际领域,而且在许多情况下,观察时间和终止时间是带信息的,同时可能会存在有偏抽样.本文在有偏抽样下,对于带信息观察时间和终止时间的纵向数据,提出了一个联合建模方法,并利用借代力量方法给出了模型参数的估计,同时获得了这些估计的渐近性质.最后,提出一种模型检验方法.Longitudinal data frequently occur in medical follow-up studies, and in many situations, there may exist informative observation times and a dependent terminal event time, as well as biased sampling. In this article, we propose joint modeling and analysis of longitudinal data with possibly informative observation times and a dependent terminal event time under biased sampling. A borrow-strength estimation procedure is developed for parameter estimation, and asymptotic properties of the proposed estimators are established. In addition, a model checking method is presented for assessing the adequacy of the model.

关 键 词：有偏抽样 带信息观察 终止时间 联合建模 潜在变量 估计方程 

分 类 号：O212.7[理学—概率论与数理统计]