On the Modified Methods for Irreducible Linear Systems with L-Matrices
Journal: Bulletin of Computational Applied Mathematics (Bull CompAMa) (Vol.6, No. 1)Publication Date: 2018-07-31
Authors : Seyyed Ahmad Edalatpanah;
Page : 119-128
Keywords : Preconditioning; accelerated overrelaxation (AOR); convergence analysis; L-matrix; directed graph;
Abstract
Milaszewicz, [Milaszewic J.P, Linear Algebra. Appl. 93,1987, 161$-$170] presented new preconditioner for linear system in order to improve the convergence rates of Jacobi and Gauss-Seidel iterative methods. Li et al.,[Li Y., Li C., Wu S., Appl. Math.Comput. 186, 2007, 379-388] applied this preconditioner and provided convergence theorem for modified AOR method. Yun and Kim [Yun J.H., Kim S.W., Appl. Math. Comput. 201, 2008, 56-64] pointed out some errors in Li et al.'s theorem and provided some correct results for convergence of the preconditioned AOR method. In this paper, we analyze their convergence properly and propose a new theorem for irreducible modified AOR method. In particular, based on directed graph, we prove that the convergence theorem of Li et al. is true, without any additional assumptions.
Other Latest Articles
- A new flexible extension of the generalized half-normal lifetime model with characterizations and regression modeling
- Fundamentals of soft category theory
- New results on the stability, integrability and boundedness in Volterra integro-differential equations
- Transitivity of parametric family of cardinality-based fuzzy similarity measures using Lukasiewicz t-norm
- Content-Based Image Retrieval Features: A Survey
Last modified: 2018-08-05 09:25:12