COURSE DESCRIPTION SysEng/CS 378 and El Eng 368 Introduction to Neural Networks and Applications

haremboingAI and Robotics

Oct 20, 2013 (3 years and 11 months ago)

116 views

COURSE DESCRIPTION





SysEng/CS 378 and El Eng 368



Introduction to Neural Networks and Applications

(Offered Spring Semesters)


Required or Elective Course:
Elective


Catalog Description:


(Lec. 3.0) Introduction to artificial neural netw
ork architectures, adaline,
madaline,back propagation, BAM, and Hopfield memory, counterpropagation
networks, self organizing

maps, adaptive resonance theory, are the topics covered.
Students experiment with the use of

artificial neural networks in enginee
ring through
semester projects. Prerequisite: Math 204 or

229 Differential Equations. (Co
-
listed
with Cp Sc 378 and Sys Eng 378)


Prerequisites by topic:
Matrix Algebra and Differential Equations


Textbooks and other required material:


M. H. Ham and I. Ko
stanic,


Principles of Neurocomputing for Science and
Engineering, (McGraw
-
Hill, NY, NY, 2001).

Duane C. Hanselman,
Bruce Littlefield
, and Bruce L. Littlefield
,
Mastering Matlab 7
, (Prentice
Hall, NJ, 2004).

MatLab and Neural Networks Tool Box (Software)


Course learning outcomes/expected performance criteria:


1.

Learn basic neural network architecture

2.

Learn basic learning algorithms

3.

Understand data pre and post
processing

4.

Learn training, verification and validation of neural network models

5.

Design Engineering applications that can learn using neural networks


Topics covered:


1.

Introduction (1 week)

2.

Network Architectures and MatLab Basics (1 week)

3.

Linear Algebr
a Review and Adaline (1 weeks)

4.

MatLab Neural Network Toolbox and Madaline (1 week)

5.

Perceptron and Learning Rules for a Single Neuron (1 week)

6.

Associative Memories (1 week)

7.

Backpropagation Learning Algorithm (3 weeks)

8.

Radial Basis Function and Self
-
Organi
zing Networks (1 week)

9.

Learning Vector Quantization and Adaptive Resonance Theory (2 weeks)

10.

Project (1 week)

11.

Reviews, Examinations (3 weeks)


Class/laboratory schedule:


One 150
-
minute lecture per week is typical MatLab is used throughout the lectures,
e
xaminations, and homework.


Contribution of course to meeting the professional component:




Students are exposed to neural network architectures and their applications in
engineering design


Relationship of course learning outcomes to ECE program outcomes:


ECE

Outcome

Course Outcomes


Comments

1

2

3

4

5

a

S

S

S

S

M

Students use optimization tools in developing
learning algorithms

b




S



c



S

S

S

Students designs engineering applications that
can adapt and learn

d







e



S

S

S


f







g





M

Students presents engineering applications

h







i







j







k

S

S

S




l








S


strong connection; M


medium connection; W


weak connection


Prepared by:

Cihan H Dagli


Date:

February 1, 2008