Operators Associated with a Pair of Nonnegative Matrices
نویسندگان
چکیده
Let Amx„, Bmxn, Xnxl, and Ymxl be matrices whose entries are nonnegative real numbers and suppose that no row of A and no column of B consists entirely of zeroes. Define the operators U, T and T by (UX)t-X? [or (UY),= Y;1], T=UB'UA and T' = UAUB'. Tis called irreducible if for no nonempty proper subset S of (1, • ■ ■ , n} it is true that X,=0, ieS; X,^0, i$ S, implies (TX),=0, ieS; (TX)i^O, i $ S. M. V. Menon proved the following Theorem. If T is irreducible, there exist row-stochastic matrices Ax and At, a positive number 6, and two diagonal matrices D and E with positive main diagonal entries such that DAE=AX and 6DBE=A'2. Since an analogous theorem holds for T', it is natural to ask if it is possible that 7" be irreducible if 7" is not. It is the intent of this paper to show that 7" is irreducible if and only if 7" is irreducible. Suppose that each of m and « is a positive integer. Let AmXn and BmXn be matrices whose entries are nonnegative real numbers and suppose that no row of A and no column of B consists entirely of zeroes. Let XnXX and YmX1 be matrices whose entries are taken from the extended real nonnegative numbers. Define the operator U by (UX)i=X{'i [or (UY)¡ = YT1] and let 0"1 = 00, co_1=0, 00+00 = 00, 000 =0, and if a>0, a 00 = 00 [1]. Define the operators T and T by T=UB'UA and T' = UAUBl where Bl is the transpose of B. Clearly . m n 1 —1 (770,= (Zbj'Za,kXk ris called irreducible if for no nonempty proper subset 5 of ^={1, •••,«} is it true that X¿=0, i e S; X^O, i $ S, implies (77Q,=0, i e S; (TX)^O, i $ S. T' is defined to be irreducible analogously. M. V. Menon [2] proved the following. Theorem 1. If T is irreducible, then there exist row-stochastic matrices Ax and A2, a positive number 0, and two diagonal matrices D and E with positive main diagonal entries such that DAE=AX and 6DBE=A2. Received by the editors March 5, 1971. AMS 1970 subject classifications. Primary 15A48; Secondary 15A15, 15A51.
منابع مشابه
On the nonnegative inverse eigenvalue problem of traditional matrices
In this paper, at first for a given set of real or complex numbers $sigma$ with nonnegative summation, we introduce some special conditions that with them there is no nonnegative tridiagonal matrix in which $sigma$ is its spectrum. In continue we present some conditions for existence such nonnegative tridiagonal matrices.
متن کاملSome Spectral Properties of an Operator Associated with a Pair of Nonnegative Matrices!1) By
An operator—in general nonlinear—associated with a pair of nonnegative matrices, is defined and some of its spectral properties studied. If the pair of matrices are a square matrix A and the identity matrix of the same order, the operator reduces to the linear operator A. The results obtained include generalizations of one of the principal conclusions of the theorem of Perron-Frobenius.
متن کاملA Modified Digital Image Watermarking Scheme Based on Nonnegative Matrix Factorization
This paper presents a modified digital image watermarking method based on nonnegative matrix factorization. Firstly, host image is factorized to the product of three nonnegative matrices. Then, the centric matrix is transferred to discrete cosine transform domain. Watermark is embedded in low frequency band of this matrix and next, the reverse of the transform is computed. Finally, watermarked ...
متن کاملFinite and infinite dimensional generalizations of Klyachko theorem
We describe the convex set of the eigenvalues of hermitian matrices which are majorized by sum of m hermitian matrices with prescribed eigenvalues. We extend our characterization to selfadjoint nonnegative (definite) compact operators on a separable Hilbert space. We give necessary and sufficient conditions on the eigenvalue sequence of a selfadjoint nonnegative compact operator of trace class ...
متن کاملA Modified Digital Image Watermarking Scheme Based on Nonnegative Matrix Factorization
This paper presents a modified digital image watermarking method based on nonnegative matrix factorization. Firstly, host image is factorized to the product of three nonnegative matrices. Then, the centric matrix is transferred to discrete cosine transform domain. Watermark is embedded in low frequency band of this matrix and next, the reverse of the transform is computed. Finally, watermarked ...
متن کامل