Print and PDF Options

COMP 4806 [0.5 credit] Numerical Linear Algebra


Matrix computations, conditioning/stability, direct methods for linear systems, classical iterative methods: Jacobi, Gauss-Seidel; modern iterative methods, Arnoldi decomposition, GMRES and other Krylov subspace-based methods for sparse and structured matrices; numerical solution of eigenvalue problems, implementation using suitable programming language, application to differential equations/optimization problems.
Also listed as MATH 4806.
Prerequisite(s): MATH 2000, (MATH 2107 or MATH 2152), MATH 3806; or permission of the School.
Lectures three hours a week.