摘要
本文介绍了在地质学中常见的多变量、多样本的矩阵数据内在的相似结构的代数矩阵研究方法.首先阐述了一般矩阵的广义奇值分解原理,以此为基础导出几种常见的标度方法,最后给出了其具体算法及应用例子.
In order to study the relationships among samples or variables in geological matrix data,the principle of scaling the data matrix to low dimension was discussed in this paper. On the basis of matric structure, we studied some scaling methods to low dimension. such as principle component-analysis, correspondence analysis. At last, algorithm of the methods was studied and example of its application was given.
出处
《物探化探计算技术》
CAS
CSCD
1993年第4期288-293,共6页
Computing Techniques For Geophysical and Geochemical Exploration
