ENEE 633: Statistical and Neural Pattern Recognition

Course Goals:

Course Prerequisite(s):

Topics Prerequisite(s):

Textbook(s)

Reference(s):

Core Topics:

1. Introduction
   1. Machine perception
   2. Pattern recognition systems
   3. Learning and adaptation
2. Bayes Decision Theory
   1. Minimu error-rate classification
   2. Classifiers, discriminant functions and decision surfaces
   3. Discriminant functions for the Normal density
   4. Error propabilities, integrals and bounds
   5. Wald’s sequential probability ratio test
3. Nonparametric Techniques
   1. Density estimation using Parzen windows
   2. Nearest-neighbor rule and error bounds
4. Linear Discriminant Functions
   1. Linear discriminant functions and decision surfaces
   2. Two-class linear separable case
   3. Perceptrons
   4. Non-separable behaviors
   5. Minimum squared-error procedures and Ho-Kashyap method
   6. Support vector machines
5. Multilayer Neural Networks
   1. Feedforward operation and classification
   2. Backpropagation algorithm
   3. Second - order methods
   4. Additional networks and training methods
6. Stochastic Methods
   1. Stochastic search
   2. Boltzman learning, networks and graphical models
   3. Evolutionary methods
7. Nonmetric methods
   1. Decision trees
   2. CART
8. Algorithm-independent Machine Learning
   1. No free lunch theorem
   2. MDL
   3. Bias and variance
   4. Resampling for estimating statistics
   5. Resampling for classifier design
   6. Estimating and comparing classifiers
   7. Combining classifiers
9. Unsupervised Learning and Clustering
   1. Mixture densities and identifiability
   2. Unsupervised Bayesian learning
   3. Data description and clustering
   4. Criterion functions for clustering
   5. Hierarchical clustering
   6. Component analysis
10. Feature Selection and Dimensionality Reduction
    1. Class separability metrics
    2. Feature selection
11. Applications to Image and Speech Recognition

Optional Topics:

Course Structure:

Grading Method: