Instructor:
Chris Atkeson
The best way to contact me is to use email:
cga@cc.gatech.edu
Office hours will be scheduled in the near future.
I am interested in numerical machine learning for
future computing environments
and
intelligent systems and robotics.
There is no TA for this course.
I communicate with you (and you all with each other) using:
- lectures
- newsgroup git.cc.class.cs4640
Desired Prerequisites:
- Calculus
- Linear Algebra
- Probability
- Statistics
Course components (estimated grading weight):
- lectures (syllabus)
- readings
- "major" assignments (45%)
- some additional "minor" assignments (5%)
- project (50%)
- no exams
Language policy: use any computer language you want. I provide sample code in C, C++, Matlab, and Java (one of these at a time, not all of them).
COLLABORATION:
You may work in groups on the assignments and project.
However each of you should turn in an independent writeup.
LATE ASSIGNMENTS:
Assignments are due by the day and time specified
in the assignment description.
Late assignments will only be accepted with prior permission of the
instructor or the TA.
Matlab documentation (from a machine with access to the COC file system)
Use file://username@lennon.cc.gatech.edu/usr/local/matlab5.2/help/helpdesk.html to get remote access to the matlab documentation. You need an account on lennon so you can log in. Replace "username" with your lennon account.
Useful documentation:
- Getting Started with MATLAB: file:/usr/local/matlab5.2/help/techdoc/basics/gettingtoc.html
- Online Manuals (in PDF), look for Neural Networks Toolbox User's Guide: file:/usr/local/matlab5.2/help/fulldocset.html
- Root of documentation tree: file:/usr/local/matlab5.2/help/docroadmap.html
Sources of Data:
Stock data
Recognize
handwritten zip codes
Reuters text
categorization