矩阵分块方法在协方差矩阵中的应用

Application of matrix partitioning method to covariance matrix

为了研究矩阵分块方法在协方差矩阵中的应用,论文介绍了协方差矩阵的基本性质及正态分布随机变量的协方差矩阵的分块特征,并结合矩阵分解和矩阵广义逆给出矩阵分块方法在协方差矩阵估计中的应用。

This paper studies the application of matrix partitioning method to covariance matrix.First,we introduced the basic properties of the covariance matrix and the partitioning characteristics of the covariance matrix of normally distributed random variables.Then with matrix decomposition and generalized inverse,we applied the method of the matrix blocking to estimate the covariance matrix.