中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊
重庆交通大学学报（自然科学版）

重庆交通大学学报（自然科学版） ›› 1993, Vol. 12 ›› Issue (3): 17-28.

• • 上一篇    下一篇

统计学中的若干矩阵研究

王松桂1, 杨虎2   

  1. 1. 中国科技大学;
    2. 重庆交通学院基础部
  • 收稿日期:1993-01-04 出版日期:1993-06-21 发布日期:2016-11-10
  • 作者简介:王松桂,男,51岁,教授
  • 基金资助:
    国家自然科学基金

Some Matrix Research in Statistics

Wang Songgui1, Yang Hu2   

  1. 1. The University of Science and Technology of China;
    2. Dept of Basic Courses
  • Received:1993-01-04 Online:1993-06-21 Published:2016-11-10

PDF

摘要： 矩阵论在现代统计的许多分支都有着广泛的应用,同时统计学又提出了许多新的有关矩阵论的研究课题,刺激了矩阵论的发展。本文旨在向读者介绍矩阵论中并与统计学密切有关的如下三方面的新发展:矩阵的{2}广义逆、矩阵的降秩逼近和偏序关系,并且讨论了这些结果的统计应用。特别是在多元分析和线性模型参数估计中的应用。

关键词: 广义逆, 偏序关系, 线性模型

Abstract: Matrix theory has come into wide use in many branches of the modern statistics, moreover their developements have been caused by many concerning problems proposed in the statistics. In this paper, three further problems related to statistics, such as {2}-generalized inverse, the singular approximating and the ordering relation of matrices are in troduced in the matrix theory. Bseides, some application of the above results are also discussed in the statistics, especially in the multivariate analysis and the parameter estimate of the linear model.

Key words: generalized inverse, ordering relation, linear model