摘要
随着科技的进步,高维数据的产生、收集变得越来越容易;传统的建模方法对变量间的结构处理较为简单,这对高维数据的分析有很大的局限性。高斯图模型可以很好的描述这种结构关系,它将变量间的结构信息通过精度矩阵(协方差矩阵的逆矩阵)中的元素来表示。本文采用CLIME方法估计精度矩阵,并且详细给出CLIME估计下的精度矩阵与其估计量在最大范数下的收敛速度的证明过程,最后将得到的精度矩阵估计量应用于股票价格上,使得股票价格之间的结构关系直观明了。
With the rapid development of technology, the generation and collection of high - dimensional data is becoming more and more easier. The traditional modeling method makes a simple handle with variable structure, which could cause significant limitation of high - dimensional data. Gaussian graph model describes the structure via the elements of the precision matrix （ the inverse matrix of covariance matrix）. This article use CLIME method to estimate precision matrix and improve the process of proof of convergence rate under norm. Finally the real exam- ple （stock price） is given. As a result, the structure of the relationship between stocks prices intuitive and easy to understand.
出处
《阴山学刊（自然科学版）》
2016年第2期16-19,共4页
Yinshan Academic Journal(Natural Science Edition)
