.po 1i
.ll 6.5i
.sz 16
.ce
\fBCS 395T: Machine Learning\fP
.sz 10
.sp
\fBWhen & Where:\fP Spring 1991, Tues. & Thurs., 9:30 - 11:00, RAS 310
.sp .5
\fBCourse Number:\fP 45265
.sp .5
\fBProfessor:\fP Ray Mooney
.sp .5
\fBOffice\fP: 4.138 Taylor, Phone: 471-9558, Email: mooney@cs.utexas.edu
.sp .5
\fBOffice Hours:\fP 10-11 Mon & Wed or by appointment
.sp .5
\fBTeaching Assistant\fP: Rick Froom, 5.148 Taylor, 471-9587, froom@cs.utexas.edu
.sp .5
\fBTA Office Hours\fP: 1-2 Mon & Wed
.sp .5
\fBPrerequisites:\fP CS 381K & CS 351 (or equivalent)
.sp .5
\fBTextbook\fP: \fIReadings in Machine Learning\fP, Shavlik & Dietterich (eds.), Morgan Kaufman, San Mateo, CA, 1990
.uh "Course Overview"
.pp
The intent of this course is to present a broad introduction to Machine
Learning in Artificial Intelligence including discussions of each of the major
approaches (see the course syllabus).  The primary focus of the course will be
on understanding the underlying algorithms used in various machine learning
systems.  Class lectures will discuss general issues in Machine Learning as
well as present abstract algorithms.  Implemented versions of many of the
algorithms will be provided in order to give a feel for how the systems
discussed in class "really work" and allow for experimentation.
.pp
Selected articles from \fIReadings in Machine Learning\fP will be read and
discussed throughout the semester.  A packet containing copies of the
transparencies I will use during class lectures (these are from last year and
may be amended somewhat during the course of the semester) is available
at the Union copy-shop near the cafeteria. I highly reccomend you buy this
packet to avoid the distraction of lots of note taking during class.
.uh "Course Requirements"
.pp
Assigned reading should be read before the corresponding class and students
should be prepared to participate in class discussion about the material.  You
will periodically be asked to hand in one page written reactions to some of the
articles.  These should be critiques or additional discussions of the ideas and
results presented in the paper rather than summaries of the material presented.
There will be about 5 homework assignments (about one every two weeks) as well
as a final project.
.pp
The homework assignments will generally be programming assignments which
involve experimenting with, extending, and/or adding features to an existing
machine learning "micro-system".  These micro-systems will be written in
Common Lisp; consequently, adequate knowledge of Lisp programming is required.
Accounts on the departmental HP workstations (2.148 and 2.144 TAY) will be
supplied for doing assignments and projects.  If students have access to other
Common Lisp machines they are welcome to use them.  Note: Be sure to hand all
assignments in on time since 25% will be deducted each day an assignment is
late.
.pp
The final project can be a more ambitious experiment or enhancement involving
existing micro-systems or an additional micro-system implementation.  In
either case, the implementation should be accompanied by a short paper
(7-10 double-spaced pages) describing the project.  Time permitting, a
class presentation can be given in lieu of a paper.  A list of suggested 
projects will be handed out later in the semester.  About half-way through the
semester you will be asked to submit a one-page project proposal.
.pp
All work in the class should be done individually and independently, including
the final project.
.uh "Grading"
.pp
The final grade will be computed as follows:
.(l
 50%  Homeworks (10% per assignment)
 35%  Final Project
 15%  Written Reactions and Class Participation
\(em\(em
100%
.)l

