ECE662 Pattern Recognition and Decision Making Processes

Fall 2004

Course Information

• Instructor:  Sarah Koskie

• Email:  skoskie@iupui.edu

• Lectures:   MW 5:45–7 pm in SL 165

• Office Hours:   T 2-4pm in SL 164F or by appointment

• Textbook: Introduction to Statistical Pattern Recognition by K. Fukunaga, Academic Press, Second edition, 1990.

• Prerequisites: ECE 302 (Probabilistic Methods in Electrical Engineering) or equivalent

• Description: Introduction to the basic concepts and various approaches of pattern recognition and decision making process. The topics include various classifier designs, evaluation of classifiability, learning machines, feature extraction and modeling.

• Tentative Outline:
  1. Introduction (Week 1)
     • A. Problems in the decision making processes
     • B. Mathematical formulation
  2. Pattern recognition and learning machines
     • Review of probability theory and linear algebra (Week 2)
     • Bayes classification (Week 3)
     • Parametric classifier design (Week 4)
     • Nonparametric design (Weeks 5, 6)
     • Estimation of classifiability (Weeks 7, 8)
     • Classifier evaluation (Week 9)
     • Learning algorithms (Week 10)
  3. Data Structure Analysis
     • Feature extraction for signal representation (Week 11)
     • Feature extraction for classification (Weeks 12, 13)
     • Clustering (Week 14)
     • Modeling and validity tests (Week 15)

• Homework Assignments    (Updated November 02, 2004)