Inference in Regression Models with Missing Covariates

缺少协变量的回归模型中的推理

• 批准号: 7896590
• 负责人: JOSEPH G IBRAHIM
• 金额: $ 20.55万
• 依托单位: UNIV OF NORTH CAROLINA CHAPEL HILL
• 依托单位国家: 美国
• 财政年份: 1997
• 资助国家: 美国
• 起止时间: 1997-09-01 至 2012-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7896590
• 关键词: Acquired Immunodeficiency Syndrome; Address; Chronic Disease; Clip; Collaborations; Computing Methodologies; Data; Data Analyses; Data Set; Development; Diagnostic; Ear; Environmental Health; Equation; Influentials; International Breast Cancer Study Group; Lasso; Linear Models; Longitudinal Studies; Malignant Neoplasms; Malignant neoplasm of lung; Malignant neoplasm of prostate; Maximum Likelihood Estimate; Measures; Methodology; Methods; Minor; Modeling; Procedures; Property; Regression Analysis; Residual state; Scheme; Statistical Models; Surveys; Testing; Veterans; Work; base; cooking; health administration; malignant breast neoplasm; novel diagnostics; response; statistics; tool

In this proposal, we propose Bayesian and frequentist methodology for local influence diagnostics and develop model assessment tools for complete data settings as well as in the presence of missing covariate and/or response data for a variety of statistical models, including generalized linear models, models for longitudinal data, and survival model. In Specific Aim 1, we develop frequentist local influence measures and goodness of fit statistics based on the general local influence development of Cook (1986), and discuss these measures for i) linear models with missing at random (MAR) and nonignorably missing covariates and ii) generalized linear models with MAR and nonignorably missing covariates. For Specific Aim 2, we develop new classes of Bayesian case influence diagnostics for the complete data setting then generalize these diagnostics to the missing data framework. The proposed methodologies in Aims 1-2 are primarily motivated from several studies in the PI’s collaborative work.

在本提案中，我们提出了用于局部影响诊断的贝叶斯和频率方法，并开发了用于完整数据设置的模型评估工具，以及用于各种统计模型（包括广义线性模型、纵向数据模型和生存模型）的协变量和/或响应数据缺失的模型评估工具。在具体目标1中，我们基于Cook（1986）的一般局部影响发展，开发了频率局部影响度量和拟合优度统计，并讨论了i)随机缺失（MAR）和不可忽略缺失协变量的线性模型和ii)随机缺失和不可忽略缺失协变量的广义线性模型的这些度量。对于具体目标2，我们为完整的数据集开发了新的贝叶斯案例影响诊断类，然后将这些诊断推广到缺失的数据框架。目标1-2中提出的方法主要来自PI合作工作中的几项研究。

项目成果

Choosing among partition models in Bayesian phylogenetics.

• DOI: 10.1093/molbev/msq224
• 发表时间: 2011-01
• 期刊: Molecular biology and evolution
• 影响因子: 10.7
• 作者: Fan Y;Wu R;Chen MH;Kuo L;Lewis PO
• 通讯作者: Lewis PO

MARM: multiscale adaptive regression models for neuroimaging data.

• DOI: 10.1007/978-3-642-02498-6_26
• 发表时间: 2009
• 期刊: Information processing in medical imaging : proceedings of the ... conference

A Comparison of Monte Carlo Methods for Computing Marginal Likelihoods of Item Response Theory Models.

计算项目响应理论模型边际似然的蒙特卡罗方法的比较。

• DOI: 10.1016/j.jkss.2019.04.001
• 发表时间: 2019
• 期刊: Journal of the Korean Statistical Society
• 影响因子: 0.6
• 作者: Liu,Yang;Hu,Guanyu;Cao,Lei;Wang,Xiaojing;Chen,Ming-Hui
• 通讯作者: Chen,Ming-Hui

Bayesian meta-experimental design: evaluating cardiovascular risk in new antidiabetic therapies to treat type 2 diabetes.

• DOI: 10.1111/j.1541-0420.2011.01679.x
• 发表时间: 2012-06
• 期刊: Biometrics
• 影响因子: 1.9
• 作者: Ibrahim JG;Chen MH;Xia HA;Liu T
• 通讯作者: Liu T

Fixed and random effects selection in mixed effects models.

混合效应模型中的固定和随机效应选择。

• DOI: 10.1111/j.1541-0420.2010.01463.x
• 发表时间: 2011-06
• 期刊: Biometrics
• 影响因子: 1.9
• 作者: Ibrahim JG;Zhu H;Garcia RI;Guo R
• 通讯作者: Guo R