Date        Topics                                                              Assignments

---

Wed 08/30 Introduction to PR systems

---

Mon 09/04 Labor Day Recess 

Wed 09/06 Introduction to PR systems

---

Mon 09/11 Bayesian decision theory

Wed 09/13 Bayesian decision theory

---

Mon 09/18 Bayesian decision theory

Wed 09/20 ML and Bayesian parameter estimation                     

---

Mon 09/25 ML and Bayesian parameter estimation                     

Wed 09/27 ML and Bayesian parameter estimation

---

Mon 10/02 ML and Bayesian parameter estimation                     Programming task assigned 

Wed 10/04 ML and Bayesian parameter estimation

---

Mon 10/09 Non parametric techniques

Wed 10/11 Non parametric techniques

---

Mon 10/16 Review Exercises                                                              

Wed 10/18 Non parametric techniques                                            Homework  assigned

---

Mon 10/23 Linear discriminant functions                                       

Wed 10/25 Linear discriminant functions                                         

---

Mon 10/30 Linear discriminant functions                                                

Wed 11/01 Unsupervised learning and clustering

---

Mon 11/06 Unsupervised learning and clustering                        Homework  due     

Wed 11/08 Unsupervised learning and clustering                       Quiz posted on WebCt

                                                                                                               (last day for quiz submission: 11/20)

---

Mon 11/13 Unsupervised learning and clustering  

Wed 11/15 Programming Presentations        

---

Mon 11/20 Programming Presentations

Wed 11/22 Neural networks     

---

Mon 11/27 Neural networks                                                               Written report due

Wed 11/29 Oral Presentations

---

Mon 12/04 Oral Presentations

---

 

OBJECTIVES

By the end of the semester, a student attending this course should:

• Understand what are the different steps that enable a Pattern Recognition (PR) system to classify an input pattern.
• Familiarize with the PR terminology such as: segmentation, feature extraction, classification, post processing, training, supervised/unsupervised learning and reinforcement learning.
• Measure or compute the minimum error rate classification and decision regions for classical density functions.
• Compute the Maximum Likelihood (ML) estimator of some density functions, distinguish between ML and Bayes methods.
• Post process classification of a sequence using Hidden Markov Models (HMM's).
• Develop code in MATLAB for nonparametric techniques such as the k-Nearest Neighbor rule to classify patterns.
• Apply linear discriminant function as a model of classification.
• Cluster data using k-means algorithm.
• Understand the Back-propagation algorithm central to the Neural Network approach.

 

This set of objectives given by the instructor is not necessarily reached by the students. Assessment should be provided to measure the distance between the outcomes and these objectives.