Computing eigenvalues: Lanczos algorithm with a new recursive partitioning method (Q1963131)
From MaRDI portal
| This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: Computing eigenvalues: Lanczos algorithm with a new recursive partitioning method |
scientific article; zbMATH DE number 1391656
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Computing eigenvalues: Lanczos algorithm with a new recursive partitioning method |
scientific article; zbMATH DE number 1391656 |
Statements
Computing eigenvalues: Lanczos algorithm with a new recursive partitioning method (English)
0 references
20 January 2000
0 references
In order to compute the eigenvalues of tridiagonal matrices obtained by applying the Lanczos algorithm to a symmetric matrix, the authors propose a recursion that combines the standard partitioning based on Sturm sequences with a local analysis that allows them to apply the Newton method when its convergence is ensured. The two numerical experiments reported suggest that a significant saving in computation time can be achieved.
0 references
Lanczos algorithm
0 references
bisection method
0 references
symmetric matrices
0 references
recursive partitioning algorithm
0 references
eigenvalues
0 references
tridiagonal matrices
0 references
Sturm sequences
0 references
Newton method
0 references
convergence
0 references
numerical experiments
0 references
0 references