THE SPECTRAL NORM OF A NONNEGATIVE MATRIX

被引:39
作者
MATHIAS, R [1 ]
机构
[1] JOHNS HOPKINS UNIV,DEPT MATH SCI,BALTIMORE,MD 21218
关键词
D O I
10.1016/0024-3795(90)90403-Y
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
Given a matrix A=[aij], define {norm of matrix}A{norm of matrix}=[{norm of matrix}aij{norm of matrix}]. Let {triple vertical-rule fence} · {triple vertical-rule fence}2 denote the spectral norm. We show that for any matrix A {triple vertical-rule fence}{norm of matrix}A{norm of matrix}{triple vertical-rule fence}2=min{r1(B)c1(C):B{ring operator}C=A} and show that under mild conditions the minimizers in (1) are essentially unique and are related to the left and right singular vectors of A in a simple way. We also show that {triple vertical-rule fence}A{triple vertical-rule fence}2≤{triple vertical-rule fence}{norm of matrix}A{norm of matrix}{triple vertical-rule fence}2 and determine the case of equality. © 1990.
引用
收藏
页码:269 / 284
页数:16
相关论文
共 11 条
[1]  
Golub G.H., 1983, MATRIX COMPUTATIONS
[2]   A SURVEY OF CONDITION NUMBER ESTIMATION FOR TRIANGULAR MATRICES [J].
HIGHAM, NJ .
SIAM REVIEW, 1987, 29 (04) :575-596
[3]  
Horn R.A, 2012, MATRIX ANAL, V2nd ed.
[4]  
HORN RA, 1989, 485 J HOPK U DEP MAT
[5]  
HORN RA, 1989, 514 J HOPK U DEP MAT
[6]  
Minc H., 1988, NONNEGATIVE MATRICES
[7]  
PANG JS, IN PRESS MATH OPER R
[8]  
PAULSEN V. I, 1986, COMPLETELY BOUNDED M
[9]  
Schur J., 1911, J REINE ANGEW MATH, V140, P1, DOI [DOI 10.1515/CRLL.1911.140.1, 10.1515/crll.1911.140.1]
[10]  
STOER J, 1970, CONVEXITY OPTIMIZATI, V1