Numerical linear algebra for sparse matrices: the Conjugate Gradient method

19.03.2019 15:00 - 15:20

Lise-Marie Imbert-Gérard (University of Maryland)

 

 

Abstract:

Sparse linear systems arise from many practical applications. Among the many methods existing to solve such systems there are two categories: direct and iterative methods.  For large sparse linear systems, unless a matrices have a very special structure, iterative methods are the method of choice. In this lecture we will introduce the concept conjugate directions and present the Conjugate Gradient algorithm.

Location:

BZ 2, 2. OG., OMP 1