有约束GM模型的几何及统计性质

Geometrical and Statistic Features of A Constrained GM Model

本文利用空间向量几何理论详细地分析了带有约束的 Gauss-Markov线性模型的几何性质,并用向量图清楚地描绘出了各向量之间的几何及统计关系,给出了各向量所具有的随机统计特性,为深刻理解该模型及有针对性的建立检验统计量提供了一种简明的数学手段。

The geometrical theory of spatial vectors is used to analyse the geo- metrical features of constrained Gauss-Markov model.The relationships among the vectors and the statistic features of themselves are graphically represented.The mathem- atical method used here is powful for people to well understand GM model theory and to establish correspondent statistics of required tests with their own purposes.