No label defined (Q4396493)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1166153
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 1166153

      Statements

      16 June 1998
      0 references
      textbook
      0 references
      linear transformations
      0 references
      pseudospectrum
      0 references
      matrices of operators
      0 references
      matrix Sylvester equation
      0 references
      algorithms
      0 references
      Euclidean spaces
      0 references
      orthogonal and unitary transformations
      0 references
      reflections
      0 references
      rotations
      0 references
      Hessenberg matrices
      0 references
      singular values
      0 references
      Schur triangularization
      0 references
      invariant subspaces
      0 references
      eigenvalues
      0 references
      Hermitian matrices
      0 references
      Hausdorff set
      0 references
      numerical range
      0 references
      linear matrix equations
      0 references
      spectrum
      0 references
      interlacing inequalities
      0 references
      Weyl inequalities
      0 references
      operator norm
      0 references
      resolvent
      0 references
      Carathéodory and Jensen inequalities
      0 references
      logarithmic subharmonicity
      0 references
      Lyapunov equations
      0 references
      spectral dichotomies
      0 references
      regular matrix pencils
      0 references
      0 references
      The book treats several important topics of the modern theory of matrices, or operators in the finite dimensional real and complex Euclidean spaces (a point of view adapted in the book). It is certainly not for beginners: a reader should be familiar with the material of the first undergraduate course in linear algebra, although formally many (but not all) necessary definitions are given in the text. The target audience will likely consist of graduate students and specialists in various fields of mathematical, physical, computational, and engineering sciences, and of teachers of linear algebra at the undergraduate and graduate levels. Various parts of the text may be successfully used in linear algebra courses, for enhancement of the standard curricula; however, the book as a whole can hardly be recommended as a text. The book consists of three unequal parts. Part 1: Introduction. Finite-dimensional real and complex Euclidean spaces are introduced axiomatically. Linear transformations, in particular, the classes of orthogonal and unitary transformations, reflections and rotations, and their matrix representations are discussed (without spectral properties, however). At an early stage, computational issues in the theory of matrices are mentioned, and the notions of the \(\varepsilon\)-spectrum (pseudospectrum) and of the spectral portrait of a matrix are introduced. The introduction concludes with an extensive treatment of Hessenberg matrices and singular values. Part 2: Matrices of Operators in the Euclidean Space. This is the central part of the book, comprising 167 pages and 8 chapters. This part contains development of many topics usually found in a typical advanced linear algebra text: Schur triangularization theorem and related material (matrix Sylvester equation, invariant subspaces); variational principles (eigenvalues as stationary values of a Hermitian form, minimax and maximin characterizations of eigenvalues of Hermitian matrices, also known as the Courant-Fischer theorem); the Hausdorff set, or the numerical range, of a matrix, including a discussion on sectorial operators; linear matrix equations and connections with invariant subspaces, projections, and dichotomies of the spectrum. Chapters 5 contains a thorough exposition, with complete and detailed proofs, of alternation theorems and their converses. These theorems, which are often referred to as interlacing inequalities, connect the eigenvalues of a Hermitian \(n\times n\) matrix and the eigenvalues of its principal \((n-1) \times (n-1)\) submatrices, or, in a different setting, the singular values of a rectangular \(m\times n\) matrix and the singular values of its \(m\times (n-1)\) submatrices. Several applications of the alternation theorems are given, including inequalities for singular values of products of matrices, and the Sturm method for calculating the number of eigenvalues of a Hermitian matrix that do not exceed a given upper bound. Chapter 6 is devoted to the Weyl inequalities and Horn theorem. These results completely describe the connections between eigenvalues and singular values of a matrix, in terms of inequalities. Again, a very detailed exposition is given, with complete proofs. As a corollary, inequalities for the singular values of the sum of two matrices are derived. Another chapter in Part 2 contains integral representation of matrix functions using the resolvent, with emphasis on projections, exponentials, and matrix powers. An estimate for the operator norm of the resolvent of an \(n\times n\) matrix \(A\) is obtained: \[ \bigl\| (I-\lambda A)^{-1} \bigr\| \leq\Biggl|\prod^n_{j=1} \bigl(1-\lambda \lambda_j (A)\bigr) \Biggr|^{-1} \cdot \prod^n_{j=2} \bigl(1+ |\lambda |\sigma_j (A) \bigr), \] where \(\lambda_j(A)\) and \(\sigma_n(A) \geq\cdots \geq\sigma_1(A)\) are the eigenvalues and the singular values of \(A\), respectively. Part 3: Applications of Spectral Analysis. The most important Algorithms. This part consists of three disjoint chapters. Chapter 12, which serves an illustrative purpose, describes matrix operators as finite dimensional models of certain partial differential operators. Applications of the theory of functions of one complex variable, notably of the Carathéodory and Jensen inequalities (which are fully proved), are found in Chapter 13. The chapter culminates in a proof of the logarithmic subharmonicity of the operator norm of the resolvent \(\| (\lambda I-A)^{-1} \|\) in any complex domain that does not contain eigenvalues of the matrix \(A\). The final chapter of the book deals with certain algorithms for computation of solutions of Lyapunov equations and of spectral dichotomies of regular matrix pencils. The book offers very limited number of problems (exercises), and most of them are of a theoretical nature. This is a significant drawback. The rather narrowly conceived bibliography comprises 33 items, of which only 5 were originally published outside of the former Soviet Union.
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references