|
Zhurnal Vychislitel'noi Matematiki i Matematicheskoi Fiziki, 2005, Volume 45, Number 3, Pages 374–382
(Mi zvmmf678)
|
|
|
|
This article is cited in 7 scientific papers (total in 7 papers)
Gradient methods with improved convergence rate
E. A. Alshinaa, A. A. Boltnevb, O. A. Kacherb a Institute of Mathematical Modeling, Russian Academy of Sciences, Miusskaya pl. 4a, Moscow, 125047, Russia
b Moscow State Institute of Electronic Technology (Technical University)
Abstract:
A very simple modification is proposed for the steepest descent and minimum residual methods that makes it possible to substantially improve the convergence rate of these methods. The proposed approach is universal and does not require any adjustment to the spectrum of a matrix. Its efficiency is confirmed by numerous computational experiments.
Key words:
iterative methods, the steepest descent method, the minimum residual method, improving the convergence rate.
Received: 26.08.2004
Citation:
E. A. Alshina, A. A. Boltnev, O. A. Kacher, “Gradient methods with improved convergence rate”, Zh. Vychisl. Mat. Mat. Fiz., 45:3 (2005), 374–382; Comput. Math. Math. Phys., 45:3 (2005), 356–365
Linking options:
https://www.mathnet.ru/eng/zvmmf678 https://www.mathnet.ru/eng/zvmmf/v45/i3/p374
|
Statistics & downloads: |
Abstract page: | 1302 | Full-text PDF : | 1296 | References: | 86 | First page: | 1 |
|