CS-644B Pattern Recognition

Lecture Descriptions, Exams, Homework and Play

Week: 1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11 - 12 - 13 - 14 - 15

Week 1-January 3

Week 2-January 8 and 10

Lecture #3:

Week 3-January 15 and 17

Lecture #5:

Week 4-January 22 and 24

Week 5-January 29 and 31

    Lecture #8:
    1. Lateral inhibition and neural networks
    2. More medial axis:
      1. Medial axis transform
      2. Medial axis pruning for noise removal
    3. Discriminant functions:
      1. Linear
        1. Threshold logic units
    Lecture #9: Tutorial on Assignments Problems
    Problem Assignment #3 ( February 28)
    1. Morphological congruence of shapes
    2. Discriminant functions
    3. Geometric probability
    4. The Bhattacharya coefficient
    Reading Assignment
    1. Handout: Medial axis and skeletal pair
    2. Handout: Mathematical foundations of learning machines
    3. Web: 9.1 - Simple classifiers
    4. Web: 9.4.4.1 - Real and artificial neurons
    5. Web: 9.4.4.2 - Threshold logic units, perceptrons and learning rules
    Suggested Play
    1. Web: 5.12.1 - Applet for medial axis of points (Voronoi diagram) in the plane
    2. Web: 5.12.2 - Applet for medial axis of points (Voronoi diagram) on the sphere
    3. Web: 12.4.1 - Applet for Rosenblatt's perceptron learning algorithm

Week 6 - February 5 and 7

Week 7 - February 12 and 14

Suggested Play

Week 8 - February 19 and 21 - Study Break

Week 9 - February 26 and 28

Week 10 - March 5 and 7

Week 11 - March 12 and 14

Week 12 - March 19 and 21

    Lecture #22:
    1. Nearest neighbor decision rules:
      1. Choosing the value of k in k-nearest neighbor rules
      2. Proximity graph nearest neighbor rules
    2. Efficient implementation of nearest neighbor decision rules:
      1. Searching nearest neighbors efficiently:
        1. Voronoi diagram methods
        1. Projection methods
        2. Space-partition trees
      1. Edited nearest neighbor rules for reduced storage requirements
        1. Hart's condenced NN-rule
        2. Optimal Voronoi editing
        3. Proximity-graph-theoretic editing methods
    Problem Assignment #5
    1. Maximum likelihood estimation of parameters
    2. The 1-Nearest-Neighbor and Bayes error rates
    3. The 2-Nearest-Neighbor decision rule
    4. The k-Nearest-Neighbor decision rule
    5. The j-th Nearest-Neighbor decision rule
    Lecture #23:
    1. Error-correction learning methods
    2. The perceptron convergence theorem
    Reading Assignment
    1. Web: 14.1.1 - Nearest neighbor decision rule tutorial
    2. Web: 14.1.4.1 - Jensen's inequality introduction
    3. Web: 14.1.4.2 - Convexity and Jensen's inequality
    4. Web: 14.2.1 - Searching nearest neighbors via projections
    5. Web: 14.3.1 - Proximity graph methods for editing nearest neighbor decision rules
    6. Web: 14.3.2 - Tutorial for proximity graph NN-rule editing
    7. Handout - Error correction learning
    8. Handout - Proof of perceptron convergence theorem
    Suggested Play
    1. Web: 14.1.1 - The nearest neighbor rule applet
    2. Web: 14.1.3 - The k-NN decision rule applet
    3. Web: 14.3.2 - Applets for proximity graph editing of NN-rules

Week 13 - March 26 and 28

    Lecture #24: 2nd Midterm Exam
    Lecture #25: Student Oral Presentations

Week 14 - April 2 and 4

Week 15 - April 9