首页> 外文学位 >Totally nonnegative matrices.
【24h】

Totally nonnegative matrices.

机译:完全非负矩阵。

获取原文
获取原文并翻译 | 示例

摘要

An m-by-n matrix A is called totally nonnegative (resp. totally positive) if the determinant of every square submatrix (i.e., minor) of A is nonnegative (resp. positive). The class of totally nonnegative matrices has been studied considerably, and this class arises in a variety of applications such as differential equations, statistics, mathematical biology, approximation theory, integral equations and combinatorics. The main purpose of this thesis is to investigate several aspects of totally nonnegative matrices such as spectral problems, determinantal inequalities, factorizations and entry-wise products. It is well-known that the eigenvalues of a totally nonnegative matrix are nonnegative. However, there are many open problems about what other properties exist for the eigenvalues of such matrices. In this thesis we extend classical results concerning the eigenvalues of a totally nonnegative matrix and prove that the positive eigenvalues of an irreducible totally nonnegative matrix are distinct. We also demonstrate various new relationships between the sizes and the number of Jordan blocks corresponding to the zero eigenvalue of an irreducible totally nonnegative matrix. These relationships are a necessary first step to characterizing all possible Jordan canonical forms of totally nonnegative matrices. Another notion investigated is determinantal inequalities among principal minors of totally nonnegative matrices. A characterization of all inequalities that hold among products of principal minors of totally nonnegative matrices up to at most 5 indices is proved, along with general conditions which guarantee when the product of two principal minors is less than another product of two principal minors. A third component of this thesis is a study of entry-wise products of totally nonnegative matrices. In particular, we consider such topics as: closure under this product, questions related to zero/non-zero patterns, and determinantal inequalities associated with this special product. Finally, a survey of classical results and recent developments, including: commonalities and differences among totally nonnegative matrices and other positivity classes of matrices; perturbations and factorizations of totally nonnegative matrices, are discussed.
机译:如果A的每个平方子矩阵(即次要)的行列式为非负(分别为正),则m×n矩阵A称为完全非负(分别为完全正)。完全非负矩阵的类别已被大量研究,并且该类别出现在多种应用中,例如微分方程,统计,数学生物学,逼近理论,积分方程和组合数学。本文的主要目的是研究完全非负矩阵的几个方面,例如频谱问题,行列式不等式,因式分解和乘积。众所周知,完全非负矩阵的特征值是非负的。但是,对于此类矩阵的特征值还存在哪些其他属性,存在许多悬而未决的问题。在本文中,我们扩展了关于一个完全非负矩阵的特征值的经典结果,并证明了一个不可约的完全非负矩阵的正特征值是不同的。我们还证明了与不可约的完全非负矩阵的零特征值相对应的Jordan块的大小和数量之间的各种新关系。这些关系是表征完全非负矩阵的所有可能约旦规范形式的必要的第一步。研究的另一个概念是完全非负矩阵的主要未成年人之间的行列式不平等。证明了最多为5个指数的完全非负矩阵的主要未成年人的产品中所有不等式的特征,以及可以保证两个主要未成年人的产品少于两个主要未成年人的产品的一般条件。本论文的第三部分是对完全非负矩阵的乘积的研究。特别是,我们考虑以下主题:该产品下的封闭,与零/非零模式有关的问题以及与此特殊产品相关的确定性不平等。最后,对经典结果和最新发展进行了调查,包括:完全非负矩阵与其他正性矩阵之间的共性和差异;讨论了完全非负矩阵的扰动和因式分解。

著录项

  • 作者

    Fallat, Shaun Michael.;

  • 作者单位

    The College of William and Mary.;

  • 授予单位 The College of William and Mary.;
  • 学科 Mathematics.
  • 学位 Ph.D.
  • 年度 1999
  • 页码 202 p.
  • 总页数 202
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 数学 ;
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号