Class Schedule
09:45 A.M. - 11:00 A. TuTh: MechE 102
Instructor
Yousef Saad
e-mail: saad at cs dot umn dot edu
Office: 5-225B EE/CS bldg -- Office Phone: (612)624-7804
Office Hours
Tu 1:00-- 2:00 PM Th 1:00 -- 2:00 PM
Course goals
This course is an introduction to sparse matrix
techniques with an emphasis on solving sparse linear systems of
equations and eigenvalue problems. The
course will cover sparse
direct methods as well as iterative methods for solving linear
systems.
About 1/3 of the class will be devoted to sparse direct methods
and 1/3 to iterative methods. The rest will be
devoted to discussed sparsity in general, eigenvalue problems,
applications, software, and
parallel implementations. There will be a few case studies and
applications to illustrate the use of sparse matrix techniques.
Another goal of the course is to consider sparsity
in contexts other
than the classical one of the solution of partial differential
equations.
For example we will explore sparsity in information
retrieval, or more generally in data analysis.
As this is an advanced class often attended by student from different
backgrounds, the emphasis of the various topics will be adjusted as the
class progresses.
This will be a practical course. Students will learn about the
algorithms and their complexity or convergence theory, and they will
also get an understanding on how to implement them and how to work
with sparse matrices in general. One of the requirements of the course
is the completion of a term project, possibly using available software
packages.