CS4252: |
Introduction to |
|
Computational Learning Theory |
|
Spring 2006 |
Instructor: Rocco Servedio
Class Manager: Andrew Wan
Email: atw12 at columbia dot edu
CONTENTS
ANNOUNCEMENTS
READING
The textbook for this course is
Introduction to Computational Learning Theory, by M. Kearns and U. Vazirani.
This book may be purchased at the Columbia Bookstore or online. Its an excellent
book, but several topics we'll cover are not in the book. Pointers to papers which will
cover these topics will be given here.
The content for the first 6 lectures will consist of the following
two papers.
The original paper by Littlestone on the Winnow algorithm can be found
here.
A survey by Avrim Blum on Online algorithms can be found
here.
A survey by Robert Schapire on Boosting can be found
here.
HOMEWORK
See instructions for submitting below.
- hw1 due Monday, Feb 13
- hw2 due Monday, Feb 27
- hw3 due Monday, Mar 13
- hw4 due Monday, Mar 27
- hw5 due Monday, Apr 10
- hw6 due Friday, April 28
- Final Project, May 10
Your problem sets must be turned in as LaTeX documents. If you're unfamiliar with
LaTeX, click here
for an introduction. All problem sets must be emailed to
Andrew Wan
by 5:00pm of the due date or they will be considered late.
Be sure that your LaTeX source code compiles correctly before you send it;
you will be penalized if your code does not compile.
Please submit both
the LaTeX source code (the file that ends with .tex) and the .ps file.