CS 786 - Machine Learning and Probabilistic Inference
Fall 2001
Department of Computer Science
University of Waterloo
Instructor: Dale Schuurmans, DC1310, x3005, dale@cs.uwaterloo.ca
Room: DC 3313
Time: Mon 2:30-5:00
Note:
Please send me email,
dale@cs.uwaterloo.ca,
if you wish to be added to the class mailing list.
This course will cover the fundamental principles of
computational learning systems. It will introduce the basic
methods used in symbolic machine learning, neural networks,
pattern recognition, and graphical probability modelling.
These techniques are now widely applied in scientific data
analysis, data mining, trainable recognition
systems, adaptive resource allocators, and adaptive
controllers. The emphasis will be on understanding the
fundamental principles that permit effective learning
in these systems, realizing their inherent
limitations, and exploring the latest advanced techniques
employed in computational learning.
Prerequisites
There are no formal prerequisites for this course---all
that is required is a basic programming capability and a
rudimentary knowledge of probability and statistics.
It would be advantageous (but not essential) to have some
prior exposure to optimization methods, statistics, and a
previous course on artificial intelligence.
Format
The course will consist primarily of prepared lectures that
cover the fundamental methods and theories of computational learning.
Depending on the interests of the class,
there may be an opportunity to cover special topics at the
end of the course.
Course work
The course grade will be based on two assignments, each worth 25%
(both of which will include a theoretical and a small programming component),
and a course project which will be worth 50% of the final grade.