Computational Methods and Data Analysis (fall 2002)
General Information
Computational Methods and Data analysis.
- Time: Fall 2002, Tuesdays 10.30-12.30 and 13.30-15.30 (w.c.)
- Location: CK N5 and A0013, respectively
- Lecturers:
- Assistant: Onno Zoeter
- Material: Syllabi
- Examination: 6 homework assignments + discussion
It is allowed to work in couples on the homework assignments. Discussion
of these assignments is per person. Final grade is at most one point
different from the average of the homework assignments.
Program
- Computational Methods
- Numerical differentiation
- Root finding
- Numerical integration
- Ordinary differential equations with initial and boundary values
- Approximation and interpolation of discrete data and functions
- Data analysis
- Probability and statistics
- interpretations, distributions, experimental design
- Parameter estimation
- maximum likelihood, least-squares (revisited)
- Nonlinear methods
- nonlinear and logistic regression, neural networks
- Exploratory data analysis
- dimension reduction, clustering, EM algorithms
Literature
- P.L. De Vries, "A first course in computational physics",
J. Wiley & Sons, New York, 1994.
- R. L. Burden, J. D. Faires, "Numerical analysis",
PWS-Kent, Boston, Mass., 1993.
- T. Hastie, R. Tibshirani, J. Friedman, "The elements of statistical
learning", Springer, New York, 2001.
MATLAB
Homework assignments as well as the final assignment can be made in any
programming language, but use of
MATLAB is highly recommended.
MATLAB is a high-level scientific and engineering programming environment
which allows easy visualization of data, has an extensive library of built-in
functions for data manipulation, and is widely used in universities and
research labs around the world.
There are many useful introductory MATLAB tutorials online. Probably
the one from the
University
of Utah will suit your needs. This new tutorial
(postscript, 1.7 Mb or
zipped postscript, 280 Kb)
is also highly recommended.
Note also that once you start MATLAB and
are within the MATLAB environment, you can simply type "doc" and a very
extensive set of online documentation will come up in your Web browser.
Type "matlab -nojvm" when you start up MATLAB for shutting down the Java
virtual machine, in case you like the simple command line interpreter
better (as we do).
Questions
Do not ever hesitate to ask us any questions if at some point you are stuck
with an assignment or need some help with the material otherwise. The
following list indicates who can probably help you best, with a backup in
case you cannot reach that person.
-
For questions about numerical methods and the corresponding work assignments,
contact Annalisa, phone
(024-36)52222 or Tom.
- For questions about data analysis and the corresponding work
assignments, contact Tom, phone
(024-36)15039 or Onno.
- For questions about MATLAB, contact
Onno, phone
(024-36)14243 or Tom.
Exercise 1
Description
MATLAB files
Exercise 2
Description
Exercise 3
Description
Data file
Exercise 4
Description
Data file
Exercise 5
Description
Data file
Exercise 6
Description
Data file
M-files