New Fast and Accurate Jacobi SVD Algorithm. II

This paper presents a new one-sided Jacobi SVD algorithm for triangular matrices computed by revealing QR factorizations. If used in the preconditioned Jacobi SVD algorithm, described in part one of this paper, it delivers superior performance leading to the currently fastest method for computing SV...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:SIAM journal on matrix analysis and applications Ročník 29; číslo 4; s. 1343 - 1362
Hlavní autoři: DRMAC, Zlatko, VESELIC, Kresimir
Médium: Journal Article
Jazyk:angličtina
Vydáno: Philadelphia, PA Society for Industrial and Applied Mathematics 01.01.2008
Témata:
Algebra
Algorithms
Applied mathematics
Decomposition
Exact sciences and technology
Linear and multilinear algebra, matrix theory
Mathematics
Nonlinear algebraic and transcendental equations
Norms
Numerical analysis
Numerical analysis. Scientific computation
Numerical linear algebra
Sciences and techniques of general use
Perturbation theory
Arithmetics
Jacobi method
Decomposition method
Eigenvalue
Implementation
15A18
Numerical analysis
Efficiency
15A12
15A23
65F22
eigenvalues; 15A09
Linear algebra
QR factorization
Floating point
65F15
Matrices
65F35
Fast algorithm
Performance
Computing method
Preconditioning
Singular value decomposition
ISSN:0895-4798, 1095-7162
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:This paper presents a new one-sided Jacobi SVD algorithm for triangular matrices computed by revealing QR factorizations. If used in the preconditioned Jacobi SVD algorithm, described in part one of this paper, it delivers superior performance leading to the currently fastest method for computing SVD decomposition with high relative accuracy. Furthermore, the efficiency of the new algorithm is comparable to the less accurate bidiagonalization-based methods. The paper also discusses underflow issues in floating point implementation and shows how to use perturbation theory to fix the imperfectness of the machine arithmetic.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
content type line 14
ISSN:0895-4798
1095-7162
DOI:10.1137/05063920X