Indefinite linear algebra and applications

From MaRDI portal





The present book is a revision and updating of a book by the three authors which was published in 1983 [Matrices and indefinite scalar products (1983; Zbl 0513.15006)]. The context of the subject is as follows. Consider the vector space \(\mathbb{C}^{n}\) with the usual inner product \(\left( ~,~\right) \). Every (nondegenerate) indefinite scalar product \(\left[ ~,~\right] \) on \(\mathbb{C}^{n}\) has the form \(\left[ x,y\right] =\left( Hx,y\right) \) where \(H\) is an invertible, nondefinite Hermitian matrix. The book is a study of the properties of \(\mathbb{C}^{n}\) and its operators with respect to such a scalar product. A large part of the book deals with the natural problem of classifying pairs \((A,H)\) (where \(A\) and \(H\) are \(n\times n\) matrices with \(H\) invertible, nondefinite Hermitian) in terms of the relation defined by: \((A_{1},H_{1} )\sim(A_{2},H_{2})\) if and only if there exists an invertible matrix \(T\) such that \(A_{1}=T^{-1}A_{2}T\) and \(H_{1}=T^{\ast}H_{2}T\). For example, an \(n\times n\) matrix \(A\) is \(H\)-selfadjoint when \([Ax,y]=[x,Ay]\) for all \(x,y\), and canonical forms for the classes of pairs \((A,H)\) where \(A\) is \(H\)-selfadjoint were already known to Weierstrass. The other main theoretical questions considered here deal with analytic functions of matrices, properties which are stable under small perturbations, and matrix polynomials \(L(\lambda)\) of the form \(A_{l}\lambda^{l}+\dots+A_{0}\) where the coefficients \(A_{i}\) are Hermitian \(n\times n\) matrices and \(A_{l}\) is invertible. For such matrix polynomials the companion matrix \[ C_{L}:=\left[ \begin{matrix} 0 & I & \dots & 0\\ 0 & 0 & \dots & 0\\ \vdots & \vdots & \ddots & \vdots\\ -\tilde{A}_{0} & -\tilde{A}_{1} & \dots & -\tilde{A}_{l-1} \end{matrix} \right] \text{ where }\tilde{A}_{i}:=A_{l}^{-1}A_{i}\text{ for each }i \] is \(H\)-selfadjoint with respect to a suitable \(nl\times nl\) Hermitian matrix \(H\) and \(H\) is nondefinite when \(n>1\), so some of the earlier results apply. The book ends with applications to analysis of the qualitative behaviour of solutions to linear differential and difference equations with constant coefficients and to the Riccati equation (\(XDX+XA+BX=C\) with \(D,A,B,C\) given rectangular matrices satisfying \(D=D^{\ast},C=C^{\ast}\) and \(A=B^{\ast}\)) which arises, for example, in optimization problems in systems theory. The book under review has a lot in common with the earlier book, although some material has been dropped, other material expanded and a few new topics added. Like the earlier book, the book is clearly written and could serve as a graduate text for students with a basic knowledge of linear algebra. Each chapter ends with a collection of straightforward exercises and a brief section on sources of the material. A substantial appendix outlines material which is used, but which might not be familiar to the student, such as the theory of perturbations of subspaces. A bibliography of 101 items lists a selection of the literature of the subject, including many books and papers to which the authors themselves have contributed.




Cited in
(only showing first 100 items - show all)








This page was built for publication: Indefinite linear algebra and applications

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5711210)