项目名称: 基于稳健估计方程的复杂纵向数据研究
项目编号: No.11501124
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 数理科学和化学
项目作者: 郑雪莹
作者单位: 复旦大学
项目金额: 18万元
中文摘要: 本课题研究复杂纵向数据的统计建模和稳健推断。主要研究内容有:1)基于稳健的广义估计方程方法研究含有缺失值的纵向数据的联合均值协方差模型的统计推断,同时也发展一些新的适用于缺失数据的有效的稳健模型和变量选择方法。此类方法的优势之一是能有效的应用于对不同个体测量时间不规则的纵向数据分析;2)针对零膨胀纵向数据,研究基于广义估计方程的统计模型和稳健推断问题;3)研究含有缺失值的零膨胀纵向数据的稳健推断。本课题的研究将通过丰富的计算机模拟检验所提方法的有效性,并将新方法应用于实际问题的分析。
中文关键词: 半参数估计;联合均值协方差模型;零膨胀数据;稳健估计;纵向数据
英文摘要: The proposal focuses on developing statistical model and robust inference procedure for complex longitudinal data. The main content is composed of: 1) Under the framework of robust generalized estimating equations, we will develop proper joint mean and covariance models for longitudinal data with missing values. Meanwhile, we will investigate some new robust joint mean and covariance models and variable selection procedure. An appealing feature of the proposed joint modeling method is that it can accommodate irregular and subject-specific observation times; 2) For zero-inflated longitudinal data set, we will study novel statistical models and robust inference based on the generalized estimating equations; 3) We will study robust inference for zero-inflated longitudinal data set with missing values. The performance of the proposed methods will be evaluated by a range of simulation studies and the developed models and methods will be applied to real data analysis.
英文关键词: semi-parametric estimation;joint mean-covariance model;zero-inflated data;robust estimating equation;longitudinal data