VECTOR MAJORIZATION VIA NONNEGATIVE DEFINITE DOUBLY STOCHASTIC MATRICES OF MAXIMUM RANK

Authors
Citation
Cs. Wong et H. Cheng, VECTOR MAJORIZATION VIA NONNEGATIVE DEFINITE DOUBLY STOCHASTIC MATRICES OF MAXIMUM RANK, Linear algebra and its applications, 261, 1997, pp. 187-194
Citations number
7
Categorie Soggetti
Mathematics,Mathematics
ISSN journal
00243795
Volume
261
Year of publication
1997
Pages
187 - 194
Database
ISI
SICI code
0024-3795(1997)261:<187:VMVNDD>2.0.ZU;2-H
Abstract
It is known that for real n-vectors x and y, y majorizes x if and only if Ay = x for some doubly stochastic matrix A of order n. Suppose tha t the coordinates of each of x and y are in nonincreasing order. Then the matrix A can be chosen to be nonnegative definite. If there is no coincidence and if the coordinates of x are not all equal, then A can be chosen to be positive definite. In this paper, we obtain a simple f ormula for calculating the maximum rank of all nonnegative definite do ubly stochastic matrices A such that Ay = x. (C) Elsevier Science Inc. , 1997.