奇异增长曲线模型中回归系数阵的最优估计

BLUE of Regression Coefficient Matrices in the Growth Curve Model with Possibly Singular Covariance

讨论了协差阵可能奇异情形下回归系数阵的最优估计,作为特例给出了协差阵非奇异时回归系数阵的最优估计。

In this paper, we give BLUE of regression coefficient matrix in the GCM with possibly singular Covariance, especially as a by-product of the result in Σ>0,meanwhile give conditions for coincidence of the BLUE and LSE.