Neural Network

Course Title: Neural Network

Course No: CSIT.416.4

Nature of the Course: Theory + Lab

Semester: 7

Full Marks: 60 + 20 + 20

Pass Marks: 24 + 10 + 10

Credit Hours: 3

Course Description

This course introduces the fundamental concepts of neural networks and essentials of artificial neural networks with single layer and Multilayer Networks. The course covers the basics and applications of neural networks, including design of neural network, learning processes, perceptron model, radial basis function and neuro-fuzzy systems.

Course Objectives

The main objective of the course is to introduce concepts of artificial neural networks. The general objectives are to: introduce the neural networks as means for computational learning, present the basic neural network architectures, give design methodologies for artificial neural networks, introduce learning theories used in neural networks, demonstrate neural network applications on real-world tasks, and explore use of fuzzy system in neural networks.

Course Contents

1. Unit I: Introduction

6 hrs

1.1. Introduction, Humans and Computers, Organization of the Brain, Biological Neuron, Biological Neuron Model, Artificial Neuron Models, Artificial Network Networks (ANN)

1.2. History of neural network research, characteristics of neural networks, Applications of ANN

2. Unit II: Basics of Artificial Neural Networks

8 hrs

2.1. Artificial Neuron Model and its Mathematical model

2.2. Activation Function, Types of Neuron Activation Function: Linear, Threshold, Sigmoid, Tangent

2.3. Models of neuron Mc Culloch –Pitts model, Perceptron, Adaline model, Madaline Model

2.4. ANN Architectures: Single-layer, Multilayer Feed Forward, Recurrent

2.5. Classification Taxonomy of ANN – Connectivity, Neural Dynamics (Activation and Synaptic)

3. Unit III: Learning Process

7 hrs

3.1. Learning, Learning Strategy (Supervised, Unsupervised, Reinforcement), Learning Rules, Types of Application

3.2. Error-Correction Learning, Memory-Based Learning, Hebbian Learning, Competitive Learning, Boltzman Learning

4. Unit IV: Single Layer Perceptrons

8 hrs

4.1. Introduction, Perceptron Models: Discrete, Continuous and Multi-Category, Training Algorithms: Discrete and Continuous Perceptron Networks

4.2. Least Mean Square Algorithm

4.3. Perceptron Convergence theorem, Limitations of the Perceptron Model, Applications

5. Unit V: Single and Multilayer Feed forward Neural Networks

8 hrs

5.1. Basic Concepts of single layered networks, Hopfield Networks

5.2. Multilayer Feed Forward Networks, Feedback Networks,

5.3. Discrete Hopfield Network

5.4. Gradient Descent, Delta Rule

5.5. Derivation of Back-propagation (BP) Training, Summary of Back-propagation Algorithm, Selection of tuning parameters in Back-propagation

6. Unit VI: Radial Basis Function Networks

5 hrs

6.1. Pattern separability and Interpolation

6.2. Regularization Theory

6.3. Regularization and Radial Basis Function (RBF) Networks

6.4. RBF network design and training

6.5. Approximation properties of RBF

7. Unit VII: Fuzzy Neural Networks

3 hrs

7.1. Neuro-fuzzy systems

7.2. Background of fuzzy sets and logic, Design of fuzzy systems

7.3. Design of fuzzy neural networks, applications of neuro-fuzzy systems

Laboratory Works

Student should have practical session for constructing artificial neural networks. The students should simulate different programs constructing neural networks for solving real world problems. The environments can be decided by the instructor, however it is highly recommended to use MATLAB, Java. The lab work should be practiced for minimum of 3 lab hours per week.

1.Artificial Neural Networks Construction and Simulation

Text Books

1.S. Haykin, Neural Networks – A Comprehensive Foundation, Prentice Hall

Reference Books

1.C. M. Bishop, Neural Networks for Pattern Recognition, Clarendon Press Oxford
2.B.Yegnanarayana, Artificial Neural Networks, Prentice Hall of India
3.Satish Kumar, Neural Networks – A Classroom Approach, Tata McGraw-Hill
4.Robert J. Schalkoff, Artificial Neural Networks, McGraw-Hill International Editions
5.Jeff Heaton , Introduction to Neural Networks for Java, Heaton Research
6.S N Sivanandam, S. Sumathi, Introduction to Neural Networks Using MATLAB , Tata McGraw-Hill

Notes:

This syllabus follows the official CSIT curriculum of Far Western University. In case of any doubt or revision, the university's published syllabus shall be considered authoritative.

Neural Networks

Neural Networks

Select University

Select Program

TabFlux . Neural Network . FWU . BSc. CSIT

Neural Network

Course Description

Course Objectives

Course Contents

Laboratory Works

Text Books

Reference Books

Notes: