Applied Math 5600
Fall 2005
MWF 10:00-10:50am, EC CR 151
Instructor: Keith Julien
Office: EC OT 235
Hours: WF 11:00-12:30
Phone: (303) 492-5753 Email:
julien@colorado.edu
Course Description:
This course will present the mathematical underpinnings
of modern numerical methods and parcticle implementations of those methods.
Motivation:
Computation has become one of the three legs of science
and engineering: Theory, Experiment and Computation. No discipline has escaped
the influence of computation and many disciplines have been reinvented because
of new computational capabilities. Today's fastest computers can perform
over 35 trillion multiplications per second. Good mathematical algorithms
are essential to effectively harness this power. This class is the first
step in understanding the mathematics of computation.
Required Text:
Kendall Atkinson, An Introduction to Numerical
Analysis, 2nd ed., Wiley, 1988.
Supplimental Text:
Golub and Van Loan, Matrix Computations, Johns
Hopkins University Press, 1899
Tests:
There will be two in-class tests and a final exam.
Programming Assignments:
There will be 4 programming assignments. The programming
language will be Matlab.
Grading
Approximatley 50% of the final grades will be determined
by the tests, 30% by the programming assignments and 20% by the homework. The homework will be spot graded.
Class participation is encouraged and will be considered
in assigning letter grades for borderline cases.
Programming
All programming will be done in MATLAB. Experience
in a programming language is a prerequisite of this course. Basic experience
with MATLAB is helpful, ut not essential. The MATLAB language is more natural
and higher-level than C or Fortran, which makes it easier to write an algorithm
directly from its mathematical description. The course textbook uses a "pseudocode"
to describe of most of the algorithms we cover. It is easy to translate
this into MATLAB.
Some helpful materials:
Matlab Quick Reference
Matlab Survival Guide
Debugging in Matlab