scholarly journals Characterization of Relationships Between the Domains of Two Linear Matrix-Valued Functions with Applications

Author(s):  
Yongge Tian

One of the typical forms of linear matrix expressions (linear matrix-valued functions) is given by $A + B_1X_1C_1 + \cdots + B_kX_kC_k$, where $X_1, \ldots, X_k$ are independent variable matrices of appropriate sizes, which include almost all matrices with unknown entries as its special cases. The domain of the matrix expression is defined to be all possible values of the matrix expressions with respect to $X_1, \ldots, X_k$. I this article, we approach some problems on the relationships between the domains of two linear matrix expressions by means of the block matrix method (BMM), the matrix rank method (MRM), and the matrix equation method (MEM). As application, we discuss some topics on the relationships among general solutions of some linear matrix equations and their reduced equations.

Symmetry ◽  
2020 ◽  
Vol 12 (11) ◽  
pp. 1791
Author(s):  
Young Hee Geum ◽  
Arjun Kumar Rathie ◽  
Hwajoon Kim

In this paper, we consider the matrix expression of convolution, and its generalized continuous form. The matrix expression of convolution is effectively applied in convolutional neural networks, and in this study, we correlate the concept of convolution in mathematics to that in convolutional neural network. Of course, convolution is a main process of deep learning, the learning method of deep neural networks, as a core technology. In addition to this, the generalized continuous form of convolution has been expressed as a new variant of Laplace-type transform that, encompasses almost all existing integral transforms. Finally, we would, in this paper, like to describe the theoretical contents as detailed as possible so that the paper may be self-contained.


2009 ◽  
Vol 16 (02) ◽  
pp. 293-308 ◽  
Author(s):  
Qingwen Wang ◽  
Guangjing Song ◽  
Xin Liu

We establish the formulas of the maximal and minimal ranks of the common solution of certain linear matrix equations A1X = C1, XB2 = C2, A3XB3 = C3 and A4XB4 = C4 over an arbitrary division ring. Corresponding results in some special cases are given. As an application, necessary and sufficient conditions for the invariance of the rank of the common solution mentioned above are presented. Some previously known results can be regarded as special cases of our results.


2003 ◽  
Vol 2003 (30) ◽  
pp. 1883-1897
Author(s):  
Nandita Rath

LetP=(pk)be a bounded positive sequence and letA=(ank)be an infinite matrix with allank≥0. For normed spacesEandEk, the matrixAgenerates the paranormed sequence spaces[A,P]∞((Ek)),[A,P]0((Ek)), and[A,P]((E)), which generalise almost all the well-known sequence spaces such asc0,c,lp,l∞, andwp. In this paper, topological duals of these paranormed sequence spaces are constructed and general representation formulae for their bounded linear functionals are obtained in some special cases of matrixA.


2021 ◽  
Vol 7 (1) ◽  
pp. 384-397
Author(s):  
Yinlan Chen ◽  
◽  
Lina Liu

<abstract><p>In this paper, we consider the common Re-nonnegative definite (Re-nnd) and Re-positive definite (Re-pd) solutions to a pair of linear matrix equations $ A_1XA_1^\ast = C_1, \ A_2XA_2^\ast = C_2 $ and present some necessary and sufficient conditions for their solvability as well as the explicit expressions for the general common Re-nnd and Re-pd solutions when the consistent conditions are satisfied.</p></abstract>


2016 ◽  
Vol 31 ◽  
pp. 465-475
Author(s):  
Jacob Van der Woude

Conditions for the existence of a common solution X for the linear matrix equations U_iXV_j 􏰁 W_{ij} for 1 \leq 􏰃 i,j \leq 􏰂 k with i\leq 􏰀 j \leq 􏰃 k, where the given matrices U_i,V_j,W_{ij} and the unknown matrix X have suitable dimensions, are derived. Verifiable necessary and sufficient solvability conditions, stated directly in terms of the given matrices and not using Kronecker products, are also presented. As an application, a version of the almost triangular decoupling problem is studied, and conditions for its solvability in transfer matrix and state space terms are presented.


2015 ◽  
Vol 30 ◽  
pp. 871-888
Author(s):  
Marcos Travaglia

This paper has been motivated by the curiosity that the circulant matrix ${\rm Circ }(1/2, -1/4, 0, \dots, 0,-1/4)$ is the $n\times n$ positive semidefinite, tridiagonal matrix $A$ of smallest Euclidean norm having the property that $Ae = 0$ and $Af = f$, where $e$ and $f$ are, respectively, the vector of all $1$s and the vector of alternating $1$ and $-1$s. It then raises the following question (minimization problem): What should be the matrix $A$ if the tridiagonal restriction is replaced by a general bandwidth $2r + 1$ ($1\leq r \leq \tfrac{n}{2 } -1$)? It is first easily shown that the solution of this problem must still be a circulant matrix. Then the determination of the first row of this circulant matrix consists in solving a least-squares problem having $\tfrac{n}{2} \, - 1$ nonnegative variables (Nonnegative Orthant) subject to $\tfrac{n}{2} - r$ linear equations. Alternatively, this problem can be viewed as the minimization of the norm of an even function vanishing at the points $|i|>r$ of the set $\left\{-\tfrac{n}{2} + 1, \dots, -1, 0, 1, \dots ,\tfrac{n}{2} \right\}$, and whose Fourier-transform is nonnegative, vanishes at zero, and assumes the value one at $\tfrac{n}{2}$. Explicit solutions are given for the special cases of $r=\tfrac{n}{2}$, $r=\tfrac{n}{2} -1$, and $r=2$. The solution for the particular case of $r=2$ can be physically interpreted as the vibrational mode of a ring-like chain of masses and springs in which the springs link both the nearest neighbors (with positive stiffness) and the next-nearest neighbors (with negative stiffness). The paper ends wiih a numerical illustration of the six cases ($1\leq r \leq 6$)corresponding to $n=12$.


2021 ◽  
Vol 6 (12) ◽  
pp. 13845-13886
Author(s):  
Yongge Tian ◽  

<abstract><p>Reverse order laws for generalized inverses of products of matrices are a class of algebraic matrix equalities that are composed of matrices and their generalized inverses, which can be used to describe the links between products of matrix and their generalized inverses and have been widely used to deal with various computational and applied problems in matrix analysis and applications. ROLs have been proposed and studied since 1950s and have thrown up many interesting but challenging problems concerning the establishment and characterization of various algebraic equalities in the theory of generalized inverses of matrices and the setting of non-commutative algebras. The aim of this paper is to provide a family of carefully thought-out research problems regarding reverse order laws for generalized inverses of a triple matrix product $ ABC $ of appropriate sizes, including the preparation of lots of useful formulas and facts on generalized inverses of matrices, presentation of known groups of results concerning nested reverse order laws for generalized inverses of the product $ AB $, and the derivation of several groups of equivalent facts regarding various nested reverse order laws and matrix equalities. The main results of the paper and their proofs are established by means of the matrix rank method, the matrix range method, and the block matrix method, so that they are easy to understand within the scope of traditional matrix algebra and can be taken as prototypes of various complicated reverse order laws for generalized inverses of products of multiple matrices.</p></abstract>


1997 ◽  
Vol 07 (02) ◽  
pp. 161-179 ◽  
Author(s):  
Michel Serfati

The present paper is devoted to some aspects of postian matrix theory over an arbitrary Post algebra, through results in Postian relative–pseudocomplementation. It is well known (for instance Rousseau [10] or Dwinger [6]) that any r-Post algebra is a Brouwerian lattice (or a Heyting algebra), that is to say, for every (a, b) in P2, the set of x in P such as a. x ≤ b admits a greatest element (a.x is inf {a, x}), called the relative inf–pseudocomplement of a in b, and denoted (b|a). In the first part of our paper, we compute the explicit expression of the disjunctive components of the relative inf–pseudocomplement (Theorem 2), which allows us to state some specific new properties of Postian pseudocomplementation, among which the cases where (b|a) is a Boolean element, and the equation (x|a) = c (Theorem 4), for which we give a consistency condition. Relative pseudo complementation in fact plays a major role in Postian structures, since we show (Theorem 5), that every element a in P is completely determined by the sequence of the (ek|a): in fact, being given (r-1) elements of P submitted to the ascending chain condition: α1 ≤ α2 ≤ … ≤ αr-1, there exists exactly one a in P such as (ek|a) = αk for every k, where the ek are the elements of the underlying chain in P. Study of pseudocomplementation properties in some Post algebra P actually helps us to enlighten relations between P and, on one hand, its center B (the set of its complemented elements, which is a Boolean algebra), on the other hand, the underlying chain. The second part is devoted to Postian linear matrix equations and inequations: in fact just like in Boolean algebras, the existence of the inf–pseudocomplement in the underlying lattice implies the residuation property is valid over the ordered semi-group of Postian matrices, equipped with the matrix product ⊗ (this was a general theorem from Blyth [3]): the Postian matrix inequation A ⊗ X ≤ B thus admits a greatest solution (Theorem 6), which is explicitly computed. This provides a consistency condition for the matrix equation A ⊗ X = B (Theorem 7). Another result states a characterization of inversible square Postian matrices. On this point of inversibility, as it is well known, the Boolean results were built in three successive steps by Wedderbrun [17], Luce [9] and Rutherford [13]. As to the Postian case, we prove in turn that a Postian matrix is inversible if and only if it is Boolean and orthogonal (Theorem 11). To prove this result, as well as Theorem 2 in the first part, we make a systematic use of the representation theorem for Post algebras by the Boolean way, as enunciated in the Preliminaries (Theorem 1). Repeated applications of the method provide a large set of conditions equivalent to the inversibility of a Postian matrix (Theorem 12). Afterwards, we examine various other Postian matrix equations and inequations, among which t A ⊗ A ≤ I, leading to the characterization of right–distributive over conjunction–Postian matrices (Theorem 9), and also the equation A ⊗ X = Ek, for which it is given a complete consistency condition (Theorem 13).


1988 ◽  
Vol 7 (2) ◽  
pp. 71-76
Author(s):  
A. Bekker ◽  
H. M. Rautenbach ◽  
J. J. J. Roux

The compatibility of pairs of matrix variate conditional densities and the uniqueness of the resulting bimatrix variate densities are investigated. Examples are given involving the matrix variate beta and Wishart conditional densities. Attention is also given to the characterization of bivector variate and bivariate distributions.


Sign in / Sign up

Export Citation Format

Share Document