MATH 497 (002) / MATH 597 (001)
Numerical Linear Algebra
Boise State University, Fall 2003
Instructor:
Stephen Brill
Office:
MG 218-A
Phone:
(208) 426-3122
Fax:
(208) 426-1356
E-mail:
brill@math.boisestate.edu
Class meetings:
3:15 p.m. to 4:30 p.m. every Tuesday and Thursday
in room MG 120.
Textbook:
Numerical Linear Algebra and Applications
by Datta.
Office hours:
Monday, Wednesday, and Friday: 12:45 p.m. to 1:30 p.m.
Tuesday and Thursday: 2:30 p.m. to 3:00 p.m.
Other times by appointment.
Academic honesty and appropriate behavior:
All students are expected to be familiar with and adhere to the policies
and standards given in the
BSU Student Code of Conduct.
In addition, if you
must have a cellular telephone or paging device on during class, please sit by the door so you can make a hasty and quiet exit if you are called.
Late work and/or extensions:
If you seek an extension on graded work
and the request occurs after the due date or time, your request will be summarily denied (except in the
most extraordinary circumstances). Such requests that occur before the due date and time will be considered on a case-by-case basis.
Contesting grades:
If you think I have graded you unfairly on a particular assignment, you must
bring this to my attention within a week from the time that the assignment
was returned to the class. After a week I will not consider any requests
to review and possibly change my grading.
Grading policy:
Your grade will be determined by your performance in three areas:
-
Homework (60%) -- Homework will be assigned on a semi-regular basis and
will consist of a combination of theoretical exercises and computer
programming. Collaborative work is not allowed.
-
Midterm Exam (20%) -- 7 October. Collaborative work is not allowed.
-
Optional extra credit problem -- In class, I provided a very detailed
derivation of the conjugate gradient (CG) method which is outlined in
Algorithm 6.10.4 on page 284. Your mission, should you choose
to accept it, is to provide the analogous derivation for the preconditioned
CG (PCG) method,
which appears on page 286 in Algorithm 6.10.5. Your derivation
must be even more detailed than mine, in that I gave you one theorem
and four lemmas without proof (which you must prove on your homework). Your
derivation of the PCG method must have everything completely justified.
If your preconditioner is named M, then you want to start with the basic CG
method, solving M^(-1)A x = M^(-1) b, and go through my derivation of the CG
method until you arrive at Algorithm 6.10.5. Your work on this project is
due at 3:15 p.m. on Thursday 11 December.
-
Final Exam (20%) -- We will discuss whether to have a take-home or
an in-class final exam. If in-class, it will be on
16 December, 3:30 p.m. to 5:30 p.m.
Your grade will be computed via the following algorithm. Let x be
the number of points accumulated throughout the semester (between 0 and
100):
A: x > 90
B: 80 < x < 90
C: 70 < x < 80
D: 60 < x < 70
F: x < 60
This page was most recently updated on 9 December 2003.
http://math.boisestate.edu/~brill/teaching/nla_f03/syll.html