ICS
663: Pattern Recognition (Fall 2009)
Day, Time and Location
MW 9:00a.m. – 10:15a.m. at POST 126
Instructor: Kyungim Baek
Office: POST 314F
Office Hours: Mondays
1p.m. – 4p.m. or by appointment
E-mail:
kyungim(AT)hawaii.edu
Tel: 808.956.8560
General Description
This course
introduces the fundamental issues of statistical pattern recognition. It
attempts to focus on general theoretical foundations of clustering,
classification and learning algorithms for pattern recognition rather than
problem-dependent details. A tentative list of topics includes Bayesian
decision theory, parameter estimation, hidden Markov models, nonparametric
techniques, supervised and unsupervised learning, linear discriminant
functions, support vector machines, and neural networks. Students are expected
to implement some algorithms using their choice of programming language and
have some background in linear algebra and probability theory.
MATH 371 or equivalent, basic knowledge of calculus, linear algebra and
probability theory, good programming skills.
Textbooks
Ÿ Required: R. O. Duda,
P. E. Hart and D. G. Stork, Pattern
Classification, 2nd Edition, John Wiley & Sons, 2001.
The text book has
a website (www.rii.ricoh.com/~stork/DHS.html)
where you can find errata lists, figures,
and etc..
Ÿ
Optional Reference: C. Bishop, Pattern Recognition and Machine Learning,
1st Edition, Springer, 2007.
Schedule (tentative)
|
Date |
Topics |
Assignment Due |
|
week1 (8/24, 26) |
Introduction, Bayesian Decision Theory |
|
|
week2 (8/31, 9/2) |
Bayesian Decision Theory |
|
|
week3 (9/9) |
ML and Bayesian Parameter Estimation |
|
|
week4 (9/14, 16) |
ML and Bayesian Parameter Estimation Component Analysis and Discriminants |
|
|
week5 (9/21, 23) |
Expectation-Maximization Hidden Markov Models |
HW 1 |
|
week6 (9/28, 30) |
Nonparametric Techniques |
|
|
week7 (10/5, 7) |
Nonparametric Techniques, Exam I |
Project Proposal |
|
week8 (10/12, 14) |
Proposal presentation |
|
|
week9 (10/19, 21) |
Linear Discriminant Functions |
HW 2 |
|
week10 (10/26, 28) |
Linear Discriminant Functions Support Vector Machines |
|
|
week11 (11/2, 4) |
Unsupervised Learning & Clustering |
|
|
week12 (11/9) |
Unsupervised Learning & Clustering |
Intermediate Report |
|
week13 (11/16, 18) |
Class Presentation |
HW 3 |
|
week14 (11/23, 25) |
Exam II, Neural Networks |
|
|
week15 (11/30, 12/2) |
Neural Networks |
|
|
week16 (12/7, 9) |
Project Presentation |
Final Project Report |
Grading
Ÿ Homework assignments: 21%
Ÿ Exams: 34%
Ÿ Final project: 45%
− Written proposal
and oral presentation (10%): short description of the topic you would like to work on and milestones for
accomplishing the project
−
Intermediate report (10%): detailed description of
the problem including background literature, description of your
approach/methods, a summary of what you have accomplished, and a revised
schedule for the rest of the semester
−
Class presentation (10%): presentation
of the major reference paper of your project
−
Final report
& presentation (15%)
Late Policy
If you hand
in late work without approval of the instructor you may receive zero credit.
Homework is
due at the beginning of class on the assigned due date.
Cooperation on Assignment
You are
allowed to discuss strategies for solving assignments with other students,
however collaboration on solutions/codes, sharing or copying of solutions/codes
is not allowed. This policy will be strictly enforced.
CHEATING WILL BE PUNISHED WITH AN F GRADE AND
REPORTED TO THE DEAN. (This policy will be equally applied to those who copy
other’s work and who allow their work to be copied.)
You may consult public literature
(books, articles, etc) for information, but you must cite each source of ideas
you adopt.
Web Page
All class
materials will be posted on the class web page at Laulima
(https://laulima.hawaii.edu/).