Fairfield University School of Engineering
Electrical and Computer Engineering

COURSE: SW461 Pattern Recognition - spring, 2017

Instructor: Jeffrey N. Denenberg

Office: Bannow 301C

Google Voice:  (203) 513-9427

Office Phone: x3330

Email:  jeffrey.denenberg@ieee.org

Web: http://doctord.webhop.net/, http://doctord.dyndhs.org/

INSTRUCTOR ASSISTANCE: TBD, and by phone or email.

CLASS HOURS: Tuesdays, 6:30 – 9:00 PM, in Bannow TBD

This programming intensive course introduces the student to the algorithms and data structures used in modern pattern recognition systems with an emphasis on those that can learn and improve their performance as they are used.  After a short review of some necessary mathematical concepts (Probability, Stochastic Processes and Vector Spaces), the student is introduced to the problem of representing real-world problems to a system.  Students will, in teams, design, build, test and report on a major subsystem of a pattern recognition system.  Selected real world applications are used to show examples of some valid representations (e.g. Speech and Handwriting) to provide insight and experience in the application of recognition systems.  Several important recognition engines are then studied and analyzed for their effectiveness as the basis for recognition, synthesis, and learning systems.  The use of additional knowledge bases dealing with the problem environment is then introduced to increase system performance.

Pre-requisites:           Basic math and algebra.  Some calculus may be helpful.
                                    Some background in probability and statistics.
                                    Programming skills (MatLab, “C” or Java)

Textbook:                   “Pattern Recognition”, Sergios Theodoridis and Konstantinos Koutroumbas,
                                    Academic Press, Ed. 4, 2009, ISBN 978-1597492720.

References:                https://www.coursera.org/learn/machine-learning/,
           
http://en.wikipedia.org/wiki/Machine_learning,
            www.sci.utah.edu/~gerig/CS6640-F2010/prob-tut.pdf
           
http://en.wikipedia.org/wiki/Variable-order_Markov_model
           
deep-learning-reinvents-the-hearing-aid

Video Lectures          Pattern Recognition:  Prof. P. S. Sastry, Indian Institute of Science, Bangalore

                        Probability Primer: MathematicalMonk on youtube

Markov Models: MathematicalMonk on youtube

Grading

Exams (2)

40%

Exercises/Participation

20%

Project

 

     Report

10%

     Presentation

10%

     Implementation

20%

Disability

If you have a documented disability and wish to discuss academic accommodations, please contact the Office of Academic Disability Support Services at (203) 254-4000, x2615, and notify the course instructor within the first two weeks of the semester.

Distance Education Students

The course is set up to support a small number distance education students.  Lecture notes and supplementary videos are accessible via links in this syllabus and via Blackboard.  You wiil submit scanned copies of any quizzes or exams via email.

Schedule:

Date

Topic

Notes

Videos

HW

1/17

Course Introduction,
Review of Probability and Vector Spaces.

Pattern Recognition Basics, Noise Introduction, Linear Algebra, PatternRecognitionTeamProjects

Introduction

Get ahead in the readings

 

Domain Representation:

 

 

 

1/24

Handwriting: A quantized, 2-D vector space in time1

Microsoft-2005 – (HMM)
NCCU-2009
NYU-2002
DoctorD’s Vector Quantizer 1987– (DTW in “C”)

 

Ch2: Computer Experiments Be ready to display and describe your solutions.
2.1, 2.7

1/31

Speech: A Multi-Dimensional vector space

Intro to ASR+HMMs
Feature Extraction
Wavelet Phoneme Recognition

 

Ch3: Computer Experiments Be ready to display and describe your solutions.
3.1, 3.4

2/7

Data Clustering

clustering algorithms-1, clustering algorithms-2, clustering algorithms-3, clustering algorithms-4

 

Ch4: Computer Experiments
4.1, 4.2, 4.3

2/14

Exam 1 – Domain Representation

 

 

 

2/21

Interim Representation Project Reports

 

 

 

 

Recognition Engines

 

 

 

2/28

Dynamic Time Warping (Dynamic Programming)

Template Matching, Dynamic Time Warp-BU

 

Ch5: Computer Experiments
4.1, 4.2

 

Markov Models

 

 

 

3/7

Nth-Order Markov Models

Smith et. al. (March 1985)
Stochastic-Modeling-of-Natural-Processes.ppt
US. #4,620,286, US. #4,599,693,
US. #4,599,692, US. #4,593,367

 

Ch6: Computer Experiments
6.1

3/14

Spring Recess – No Class

 

 

 

3/21

HMM
Hidden Markov Models

hmm-buffalo

 

Ch7: Computer Experiments 7.1, 7.2

3/28

Neural Networks

deep-learning-reinvents-the-hearing-aid

 

 

4/4

Higher Order Knowledge

 

 

 

 

Exam 2 – Recognition Engines

 

 

Ch8: Computer Experiments
8.1

4/11

Interim Recognition Project Reports

 

 

Ch9: Computer Experiments
9.1, 9.3

 

Team Project Presentations

 

 

 

4/18

Representation 1

 

 

 

4/25

Representation 2

 

 

 

5/2

Recognition Engine 1

 

 

 

5/9

Recognition Engine 2

5/4-5/11 Final Exam Week

 

 

 


 

Learning Outcomes

No.

Outcome

Cognitive Level

ABET a-k

1

The student will understand the problems encountered, approaches to and how to design a pattern recognition system.

Knowledge, Application,

& Synthesis

a, c, e, f, h, j, k

2

The student will understand how to test a recognition system and quantify the results.

Analysis

a, e, k

3

The student will be able to use mathematical analysis to reducing the complexity of real world pattern recognition applications.

Application

a, c, e, k

4

The student will be able to design and implement a major component of a pattern recognition system.

Synthesis

a, c, e, k

 

Teacher Responsibilities

Distribute syllabus.

Review the material described in the syllabus.

Explain material.

Identify alternate reading assignments or books that clarify the material.

Relate material to "real world" situations when possible.

Answer questions.

Be available to discuss problems.

Be receptive to new ideas.

Announce business/class conflicts in advance.

Make up missed classes, but classes are never cancelled as they will be done on-line.

Prepare and administer exams.

Grade fairly.

Assign appropriate home problems.

Homework policy – reviewed in class, Quizzes cover the HW topics

Student Responsibilities

Be familiar with the prerequisite material as well as the Computer Tools and Tutorials:

Ask questions and stay current.

Study the material described in the syllabus. Preferably before it covered in class and do some of the problems with answers in the back of each assigned chapter.

Complete the assigned homework.

Obtain class notes and homework if a class is missed.  View Author’s lecture video on that week’s topic(s)

Use the library and the Internet to obtain supplemental material.

Prepare for quizzes/exams.

Ask for help from me (I have office hours) and/or your fellow students.

Note: All exams in this course are open book, but not open computer (or phone) so relying on an eBook or PDF will put you at a disadvantage.