青岛大学学报（自然科学版）2018，Vol.31Issue(1)：10-14,5.DOI:10.3969/j.issn.1006-1037.2017.11.03
基于图的高斯分布的性质研究
Study on the Properties of Gauss Distribution Based on Graph
摘要
Abstract
The gaussian graphical model is a graphical model based on the assumption of the gauss distribution,which provides a powerful framework for multivariate statistical modeling.We first take the random gaussian graphical model with four nodes as an example,and then extend it to the gaussian graphical model whose number of nodes is greater than four.By using the method of inversion of the precision matrix,the covariance matrix of the model is obtained.The changes of variance,covariance and correlation coefficient between two random variables are discussed after removing one edge randomly in the gaussian graphical model.The conclusion that the gaussian graphical model with more edges or more nodes has a smaller variance is got.关键词
高斯图模型/精度矩阵/随机图模型Key words
gaussian graphical model/precision matrix/random graphical model分类
数理科学引用本文复制引用
韩仃仃,许成,逄潇..基于图的高斯分布的性质研究[J].青岛大学学报（自然科学版）,2018,31(1):10-14,5.基金项目
国家自然科学基金(批准号:60664026)资助. （批准号:60664026）
