For a given n x n matrix the ratio between the componentwise distance to the nearest singular matrix and the inverse of the optimal BauerSkeel condition number cannot be larger than (3 + 2&)n. In this note a symmetric matrix is presented where the described ratio is equal to n for the choice of most
Block matrices and symmetric perturbations
β Scribed by Alicja Smoktunowicz
- Publisher
- Elsevier Science
- Year
- 2008
- Tongue
- English
- Weight
- 120 KB
- Volume
- 429
- Category
- Article
- ISSN
- 0024-3795
No coin nor oath required. For personal study only.
β¦ Synopsis
We prove that if A = [A ij ] β R N,N is a block symmetric matrix and y is a solution of a nearby linear system (A + E)y = b, then there exists F = F T such that y solves a nearby symmetric system (A + F )y = b, if A is symmetric positive definite or the matricial norm ΞΌ(A) = ( A ij 2 ) is diagonally dominant. Our blockwise analysis extends existing normwise and componentwise results on preserving symmetric perturbations (cf. [
π SIMILAR VOLUMES
This paper addresses the finest block triangularization of nonsingular skewsymmetric matrices by simultaneous permutations of rows and columns. Hierarchical relations among components are represented in terms of signed posets. The finest block-triangular form can be computed efficiently with the aid
Guo [W. Guo, Eigenvalues of nonnegative matrices, Linear Algebra Appl. 266 (1997) 261-270] sets the question: if the list Ξ = {Ξ» 1 , Ξ» 2 , . . . , Ξ» n } is symmetrically realizable (that is, Ξ is the spectrum of a symmetric nonnegative matrix), and t > 0, whether or not the list Ξ t = {Ξ» 1 + t, Ξ» 2
The simultaneous diagonalization of two real symmetric (r.s.) matrices has long been of interest. This subject is generalized here to the following problem (this question was raised by Dr. Olga Taussky-Todd, my thesis advisor at the California Institute of Technology) : What is the first simultaneou
We obtain eigenvalue perturbation results for a factorised Hermitian matrix H = GJ G \* where J 2 = I and G has full row rank and is perturbed into G + Ξ΄G, where Ξ΄G is small with respect to G. This complements the earlier results on the easier case of G with full column rank. Applied to square facto