Mathematical Sciences: Order Restricted Inference in Multivariate Variance Components Models

数学科学：多元方差分量模型中的阶数限制推理

• 批准号: 9003467
• 负责人: Richard Dykstra
• 金额: $ 3.01万
• 依托单位: University of Iowa
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1990
• 资助国家: 美国
• 起止时间: 1990-06-01 至 1992-11-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9003467&HistoricalAwards=false
• 关键词: Mathematical; Sciences; Order; Restricted; Inference

The research involves the study of estimation in unbalanced multivariate variance components models using an approach based on a Fenchel duality formulation of the problem. The effort will be concentrated in estimation and testing of covariance matrices in multivariate variance component models, relating the results to appropriate fields of statistics, and studying rates of convergence of the proposed algorithms. The multivariate random effects model involves constraints that the covariance matrices of the random effects be nonnegative definite which leads to a Lowner ordering of matrices. Algorithms will be investigated find to the restricted maximum likelihood estimators of such matrices.

该研究涉及非平衡多变量方差分量模型估计的研究，该方法基于问题的Fenchel对偶形式。这项工作将集中在多变量方差分量模型中协方差矩阵的估计和测试，将结果与适当的统计学领域联系起来，并研究所提出的算法的收敛速度。多元随机效应模型的约束条件是随机效应的协方差矩阵是非负定的，这导致了矩阵的Lowner序。将研究寻找此类矩阵的受限最大似然估计的算法。