Description

Overview of elementary neurophysiology and the gross simulation of the network of biological neurons in the central nervous system into an artificial neural network (ANN). Study of various neural network architectures such as multilayered feedforward artificial networks (Adaline, Madaline, perceptron), backpropagation and counterpropagation networks, bidirectional associative memories, Kohonen’s self-organizing maps and recurrent networks (Hopfield networks, Boltzman machines). Other important topics covered are Adaptive Resonance Theory (ART 1, 2, 3), spatiotemporal pattern classification and the application of ANNs to various disciplines such as medicine, pattern recognition and robotics.

Prerequisites:

• CSC 305 + Lab : Object-Oriented Programming

• MAT 203 : Linear Algebra: Vectors and Matrices

• MAT 212 : Probability & Statistics for Science and Engineering

Expected Outcome(s):

• Describe the relation between real brains and simple artificial neural network models.
• Explain and contrast the most common architectures and learning algorithms for multilayer perceptrons, radial-basis function networks, committee machines and Kohonen self-organizing maps.
• Discuss the main factors involved in achieving good learning and generalization performance in neural network systems.
• Identify the main implementation issues for common neural network systems.
• Evaluate the practical considerations in applying neural networks to real classification and regression problems.

Suggested Books:

1. Neural Network Design by Martin T. Hagan
2. Neural Networks by Simon Haykin
3. Neural Networks for Pattern Recognition by C. Bishop

Grading Policy:

Biweekly Quiz, One Midterm Exam, One Final Exam, Project

Letter Grade	Marks	Grade Point
A	90 - 100	4.00
A-	85 - 89	3.70
B+	80 - 84	3.30
B	75 - 79	3.00
B-	70 - 74	2.70
C+	65 - 69	2.30
C	60 - 64	2.00
C-	55 - 59	1.70
D+	50 - 54	1.30
D	45 - 49	1.00
F	00 - 44	0.00

---