CSC 475/592
Spring 2004
CSC 475/592 is an introduction to the methods and theory of pattern recognition. The course will begin by formulating pattern recognition problems from diverse areas such as computer network traffic monitoring, typewritten character identification, and acoustic signal processing as well as digital processing of insect specimen images, military targets, or human faces subject to the effects of varying position or aging. Basic techniques such as conventional statistical approaches, neural networks, fuzzy logic systems, support vector machines, or (linear) principal component analyses (PCA) will be presented. In addition, the course will present advanced tools for non-linear PCA, clustering, and feature extraction. Student teams will be expected to implement selected pattern recognition methods using a high-level programming language, apply those methods to selected problem domains, and document the findings in written reports.
The course meets Mondays and Wednesdays from
Richard O. Duda, Peter E. Hart, and David G. Stork, Pattern Classification, John Wiley & Sons, Inc., New York, NY, 2001, ISBN 0-471-05669-3.
Since this is a research-oriented course, the grading scheme will reflect a strong emphasis on conducting studies and reporting the results. Accordingly, grading for this course will be based upon performance on a mid-term examination (20%), the final examination (30%), and both the conduct and presentation of research on a pattern recognition/classification problem (50%). Notice again that the research component is a central and sizeable portion of the grade that must be taken very seriously.
Incomplete grades are
given only very rarely and only when the student is
-
otherwise passing the course,
-
able to complete the work of the course entirely on his/her own, and
-
prevented from completing the course by verified unforeseen circumstances
beyond the control of the student.
The instructor MUST be
able to certify all three of these
factors to the chair before assigning a grade of "I".
Regular attendance and vigorous participation in class are expected but not required. However, if you desire the "benefit of the doubt" in any matter related to your grade in the class, you will routinely be present, ask relevant questions, and cooperate with the instructor as well as the course objectives. Each student is personally responsible for material covered during each class meeting.
Research questions will be posed early in the semester and class members will be expected to embrace the opportunity to conduct research. Research reports are expected and should include findings organized as follows:
Gene A. Tagliarini, PhD
Associate Professor of Computer Science
Bear Hall, Room 122
962-7572
T-R,
M-W,
Other office hours are readily available by appointment.
tagliarinig@uncw.edu
If you have a disability and need reasonable accommodation in this course, you should inform the instructor of this fact in writing within the first week of class or as soon as possible. If you have not already done so, you must register with the Office of Disability Services in Westside Hall (ext. 3746) and obtain a copy of your Accommodation Letter. You should then meet with your instructor to make mutually agreeable arrangements based on the recommendations of the Accommodation Letter.