论文信息 - More on Extremal Positive Semidefinite Doubly Stochastic Matrices

More on Extremal Positive Semidefinite Doubly Stochastic Matrices

Abstract Let K n be the convex set of n × n positive semidefinite doubly stochastic matrices. We show that for matrices with a special type of graph extremality can be determined by graph and rank. We also give a complete classification of the extreme matrices in K 5 , and in particular give examples of two matrices in K 5 with the same graph and rank, one extreme and one not.

Abraham Berman | Naomi Shaked-Monderer | A. Berman | Naomi Shaked-Monderer

[1] Positive definite doubly stochastic matrices and extreme points , 1986 .

[2] R. Grone,et al. Nonchordal positive semidefinite stochastic matrices , 1992 .

[3] Bernard Ycart. Extrémales du cône des matrices de type non négatif, à coefficients positifs ou nuls , 1982 .

[4] A. Berman,et al. Bipartite completely positive matrices , 1988, Mathematical Proceedings of the Cambridge Philosophical Society.

[5] S. Pierce,et al. Extremal positive semidefinite doubly stochastic matrices , 1991 .

[6] Leiba Rodman,et al. The ranks of extremal positive semidefinite matrices with given sparsity pattern , 1989 .

[7] R. Brualdi. The DAD theorem for arbitrary row sums , 1974 .