CPSC 517: Sparse Matrix Computations, Term 1, 2013/2014
CPSC 517: Sparse Matrix Computations
Term 1, 2013/2014
Tentative Course Outline
Sparse matrices
- relevant applications: PDEs, constrained optimization, social networks,...
- data storage schemes
Linear systems
- stationary methods: Jacobi, Gauss-Seidel, SOR
- Krylov subspace solvers: minimum residual methods,
conjugate gradients and bi-conjugate gradients
- preconditioning techniques
- multigrid
- direct solvers: matrix graphs, fill-in reducing orderings,
bandwidth reduction
Least squares and constrained optimization problems
- iterative solution methods for least squares: CGLS, LSQR
- techniques for solving saddle-point systems
Eigenvalue problems
- power method and orthogonal iteration
- Lanczos and Arnoldi
- Jacobi-Davidson
Other topics, time permitting
Back to CPSC 517
Webpage