English Türkçe için tıklayın
Chairman's Welcome Message Information
 Lisans Öğrencileri Lisansüstü Adayları



Credits: 3


Catalog Description: Neural Networks                    (3+0+0) 3
Principals of neural computing.  Architectural analysis of different neural network models (Hopfield Model, Single Perceptron, Multilayer Perceptron etc.).  Learning algorithms.  Back propogation algorithm and local minima problem.  Dynamics of recurrent neural networks.  Applications of neural networks for control systems, system identification, associative memories, optimization problems etc..  Computer simulation homeworks and final project. 


Coordinator: M. Kemal Cılız, Assoc. Professor of Electrical Engineering


Goals: This course aims to expose the students to the basic principles of Artificial Neural Networks (ANNs). The main objective is to introduce the widely accepted ANN models and have the students learn these models by simulation exercises.


Learning Objectives:

At the end of this course, students will be able to:

Understand what artificial neural networks are.
Understand  the fundamental concepts of feedforward and feedback  artificial neural network models used for various engineering applications.
Design and Build artificial neural network models and test their performances.


Textbook: Class notes will be used. There is no designated textbook for the course.


Reference Text:

H. Hertz, A. Krogh and R.G. Palmer, Introduction to the Theory of Neural Computation, Redwood City, CA, Addison Wesley, 1991.


Prerequisites by Topic:

Linear algebra
Basic system theory concepts
Ordinary differential equations
Good knowledge of programming techniques



 Introduction and a historical perspective. General Features of Neural Networks,Learning, Delta-Rule, Classification of Neural Networks. Computational Intelligence. (1 Week)
 Recurrent Networks (Hopfield Model), Dynamics of Binary and Continuous Recurrent Networks. (2 Weeks)
 Feedforward Networks. Single Layer Perceptron, LMS Rule, Multilayer Perceptron. (3 Weeks)
 Backpropagation Learning Algorithm. Derivation, problems associated with  backpropagation learning. (2 Weeks)
 Unsupervised learning. Competitive learning concepts. (1 Week)
 Well known engineering applications of recurrent and feedforward neural networks (based on the interest of the class). (3 Weeks)
 New developments in NN research. Review of Recent Journal Articles on the Theory and Applications of Neural  Networks.(1 Week)

Course Structure: The class meets for three lectures a week, each consisting of  50-minute sessions. 4 sets of homework problems are assigned per semester.  Each homework assignment requires computer programming skills. There is a final exam.


Computer Resources: Students are encouraged to use C programming language for their homework assignments.


Laboratory Resources: None.



Homework sets (60%)
A final exam (40%).


Outcome Coverage:

(a) Apply math, science and engineering knowledge. This course covers the basic fundamentals on Artificial Neural Networks. It requires basic linear algebra, system modelling, and simple optimization techniques. Using these mathematical tools, students build ANN models to be used in certain engineering applications.
(e) Ability to identify formulate and solve engineering problems. The course teaches the basic ANN models and tools to be used in various engineering problems. Hence students have to formulate the engineering problems in a way to apply ANN s as a solution tool.
(k) Use of modern engineering tools.  The course requires usage of computer simulation tools and graphical visualization programs. Students are required to simulate all the ANN models discussed in the class and apply these models to certain engineering problems.
Prepared By:
Kemal Cılız


Boğaziçi Üniversitesi - Elektrik ve Elektronik Mühendisliği Bölümü

34342 - Bebek / İSTANBUL

Tel: +90 212 359 64 14
Fax: +90 212 287 24 65







Copyright 2009 by Bogazici University EE Dept.    |   Privacy Statement   |   Terms Of Use     |     Login