摘要
用Stieltjes变换给出一般高维样本协方差矩阵的极限密度函数的显示表达式,包括:样本元素独立且均值为0,方差为常数的样本协方差矩阵;一个样本协方差矩阵与单位阵的和;样本元素方差不等但只取两值的样本协方差矩阵;两个不同的样本协方差矩阵之和.
Using the method of Stieltjes transform,we gave the explicit expressions of the limiting spectral density functions of the general high dimensional sample covariance matrices,which included:the sample covariance matrix whose elements were independent with zero mean and constant variance;the sum of a sample covariance matrix and a unit matrix;the sample covariance matrix which satisfied that the variance of the elements were different but only had two values;the sum of two different sample covariance matrices.
出处
《吉林大学学报(理学版)》
CAS
CSCD
北大核心
2016年第3期499-505,共7页
Journal of Jilin University:Science Edition
基金
国家自然科学基金(批准号:11301063)
吉林省教育厅"十二五"科学技术研究项目(批准号:吉教科合字[2015]第52号)
长春理工大学科技创新基金(批准号:XJJLG-2014-01)
关键词
随机矩阵
极限谱密度函数
样本协方差矩阵
random matrix
limiting spectral density(LSD)function
sample covariance matrix