List of Figures

LIST OF FIGURES
FIGURE No.
NAME OF FIGURE
PAGE No.
2.1
Components of neuron
12
2.2
The synapse
12
2.3
Neuron model
13
2.4
A simple neuron
14
2.5
MCP neuron
16
2.6
Network layers
17
2.7
Feed forward networks
18
2.8
Feed back networks
19
2.9
The Mcculloh Pitts model
19
2.10
Transfer functions
22
2.11
Back Propagation network
24
2.12
The principle layout for a reinforcement
26
learning agent
2.13
Flowchart for implementation of applications
28
3.1
History of VHDL
34
4.1
Back Propagation network
45
5.1
Circuit architecture of prototype chip
51
5.2
Block diagram of neuron
52
5.3
Block diagram of synapse
53
5.4
Block diagram of error generator
54
FIGURE No.
PAGE No.
NAME OF FIGURE
5.5
Block diagram of Back Propagation
55
6.1
Simulation results for synapse
61
6.2
Simulation results for neuron
63
6.3
Simulation results for synapse
65
6.4
Simulation results for neuron
67
6.5
Simulation results for error generator at
69
output
6.6
Simulation results for weight update unit
71
6.7
Simulation results for weight transfer unit
73
6.8
Simulation results for error generator at input
75
6.9
Simulation results for weight update unit
77
6.10
Simulation results for weight transfer unit
79
6.11
Simulation results for final entity
82-86
LIST OF TABLES
TABLE No.
NAME OF TABLE
PAGE No.
2.1
Truth Table before applying the firing rules
15
2.2
Truth Table after applying the firing rules
15
5.1
Table of signals for Back Propagation
56
algorithm
TABLE OF CONTENTS
CHAPTER TITLE
PAGE No.
Certificate .i
Acknowledgementii
Abstract...iii
Table
List
of
of
Contentsiv
Figures.ix
List of Tables..xi
Chapter 1 Introduction
1.1 Introduction....................................................................................................1
1.2 Advantages of neural networks .1

1.3 Learning in neural networks...............................................2
1.4 Overview of back propagation algorithm...2

1.5 Use of back propagation neural network solution.. ...3
1.6 Objective of thesis......5
1.7 Organization of thesis.5
Chapter 2 Introduction to neural networks

2.1 Introduction ...................................................................................................6
2.2 History of Neural Networks...7

2.3 Advantages of Neural Networks....9
2.4 Neural Networks versus Conventional Computers..10
2.5 Human and Artificial Neurons-Investigating the Similarities......11
2.5.1 Learning process in Human Brain. .11

2.5.2 Human Neurons to Artificial Neurons....13
2.6 An engineering approach..13
2.6.1 A simple neuron .........13
2.6.2 Firing rules..............................................14
CHAPTER TITLE
PAGE No.
2.6.3 A more complicated neuron........16

2.7 Architecture of Neural Networks.17
2.7.1 Network Layers...17

2.7.2 Feed-Forward Networks..18
2.7.3 Feedback Networks.18
2.8 Perceptrons...19
2.8.1The Learning Process...20
2.8.2 Transfer Function....22

2.9 Learning Algorithms of Neural Networks....23
2.9.1 Supervised Learning....24
2.9.1.1 The Back Propagation Learning..24
2.9.1.2 Reinforcement Learning..25
2.9.2 Unsupervised Learning26

2.9.2.1 Hebbian Learning26
vi
2.9.2.2 Competitive Learning...27

2.10Applications of Neural Networks...27
2.10.1 Neural Networks in Practice.28
2.10.2 Neural Networks in Medicine..29

2.10.2.1 Modeling and Diagnosing the Cardiovascular System .29
2.10.2.2 Electronic Noses.30
2.10.2.3 Instant Physician.30
2.10.3 Neural Networks in Business...31
Chapter 3 VHSIC Hardware Description Language (VHDL)
3.1 Introduction.....................................................................................32
3.2 History of VHDL.33
3.3 Advantages of VHDL...34
3.4 Entities and Architecture..35
3.5 Data Types ...35
3.6 Design Units.36
CHAPTER TITLE
PAGE No.
3.6.1 Entities.....37
3.6.2 Architectures...37
3.6.3 Packages and Package Bodies.38
3.6.4 Configurations.39
3.7 Levels of abstraction...39
3.7.1 Behavior.....40
3.7.2 Dataflow.41
3.7.3 Structure.41
3.8 Objects, Data Types and Operators..42
3.8.1 Using Signals...42
3.8.2 Using Variables...42
vi
i
Chapter 4 Back propagation Algorithm

4.1 Introduction..43
4.2 History of algorithm.43
4.3 Leaning with the back propagation algorithm..44
4.4 Implementation of back propagation algorithm...46
4.5 Drawbacks....48
Chapter 5 Hardware Implementation
5.1 Introduction..50
5.2 Chip architecture..51
5.3 Detailed analysis of the block diagram and hardware implementation54
5.4 Brief Description of the Entities...56

5.4.1 Synapse56
5.4.2 Neuron.57
5.4.3 Error Generator at the output...57
5.4.4 Weight update..58
5.4.5 Weight transfer unit.58
CHAPTER TITLE
PAGE No.
5.4.6 Error Generator at the Input58
Chapter 6 Simulation Results

6.1 Synapse.60
6.1.1 Inputs and outputs for the entity Synapse...60
6.1.2 Simulation Results for the entity Synapse...61
6.2 Neuron..62
6.2.1 Inputs and outputs for the entity Neuron.62
6.2.2 Simulation Results for the entity Neuron63

6.3 Synapse.64
vii
i
6.3.1 Inputs and outputs for the entity Synapse...64

6.3.2 Simulation Results For the entity Synapse..65
6.4 Neuron..66
6.4.1 Inputs and outputs for the entity Neuron.66
6.4.2 Simulation Results for the entity Neuron67

6.5 Error Generator at Output.68
6.5.1 Inputs and outputs for the entity Error Generator at the output..68
6.5.2 Simulation Results for the entity Error Generator at the output..69
6.6 Weight Update..70

6.6.1 Inputs and outputs for the entity Weight Update............................70
6.6.2 Simulation Results for the entity Weight Update71
6.7 Weight Transfer72

6.7.1 Inputs and outputs for the entity Weight Transfer..........................72
6.7.2 Simulation Results for the entity Weight Transfer..73
6.8 Error Generator at Input...74

6.8.1 Inputs and outputs for the entity Error Generator at Input.74
6.8.2 Simulation Results for the entity Error Generator at Input75
6.9 Weight Update.76

6.9.1 Inputs and outputs for the entity Weight Update....................76
6.9.2 Simulation Results for the entity Weight Update ...77
6.10 Weight Transfer..78
CHAPTER TITLE
PAGE No.
6.10.1 Inputs and outputs for the entity Weight Transfer............................78

6.10.2 Simulation Results for the entity Weight Transfer ...79
6.11 Final Entity.80

6.11.1 Inputs and Outputs for the Final Entity.80
6.11.2 Simulation Results for the Final Entity.82
Chapter 7 Conclusion and Future Scope
ix
7.1 Conclusion..8
8
7.2 Future scope...89
References .....90
List of
Publications
.9
3
USES:
One of the most common applications of NNs is in image processing. Some examples would be:
identifying hand-written characters; matching a photograph of a person's face with a different photo
in a database; performing data compression on an image with minimal loss of content. Other
applications could be voice recognition; RADAR signature analysis; stock market prediction. All of
these problems involve large amounts of data, and complex relationships between the different
parameters.
It is important to remember that with a NN solution, you do not have to understand the solution at all.
This is a major advantage of NN approaches. With more traditional techniques, you must understand the
inputs, and the algorithms, and the outputs in great detail, to have any hope of implementing something
that works. With a NN, you simply show it: "this is the correct output, given this input". With an
adequate amount of training, the network will mimic the function that you are demonstrating.
Further, with a NN, it is ok to apply some inputs that turn out to be irrelevant to the solution - during
the training process; the network will learn to ignore any inputs that don't contribute to the output.
Conversely, if you leave out some critical inputs, then you will find out because the network will fail
to converge on a solution

List of Figures

Uploaded by

Copyright:

Available Formats

List of Figures

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

List of Figures

Uploaded by

Copyright:

Available Formats

LIST OF FIGURES

Feed forward networks

Feed back networks

The Mcculloh Pitts model

Back Propagation network

The principle layout for a reinforcement

Flowchart for implementation of applications

Back Propagation network

Circuit architecture of prototype chip

Block diagram of neuron

Block diagram of synapse

Block diagram of error generator

Block diagram of Back Propagation

Simulation results for synapse

Simulation results for neuron

Simulation results for synapse

Simulation results for neuron

Simulation results for error generator at

Simulation results for weight update unit

Simulation results for weight transfer unit

Simulation results for error generator at input

Simulation results for weight update unit

Simulation results for weight transfer unit

Simulation results for final entity

Truth Table before applying the firing rules

Truth Table after applying the firing rules

Table of signals for Back Propagation

1.2 Advantages of neural networks .1

1.4 Overview of back propagation algorithm...2

1.6 Objective of thesis......5

1.7 Organization of thesis.5

Chapter 2 Introduction to neural networks

2.2 History of Neural Networks...7

2.5.1 Learning process in Human Brain. .11

2.6.3 A more complicated neuron........16

2.7.1 Network Layers...17

2.8.2 Transfer Function....22

2.9.2 Unsupervised Learning26

2.9.2.2 Competitive Learning...27

2.10.2 Neural Networks in Medicine..29

Chapter 3 VHSIC Hardware Description Language (VHDL)

Chapter 4 Back propagation Algorithm

Chapter 5 Hardware Implementation

5.4 Brief Description of the Entities...56

5.4.5 Weight transfer unit.58

5.4.6 Error Generator at the Input58

Chapter 6 Simulation Results

6.2.2 Simulation Results for the entity Neuron63

6.3.1 Inputs and outputs for the entity Synapse...64

6.4.2 Simulation Results for the entity Neuron67

6.6 Weight Update..70

6.7 Weight Transfer72

6.8 Error Generator at Input...74

6.9 Weight Update.76

6.10 Weight Transfer..78

6.10.1 Inputs and outputs for the entity Weight Transfer............................78

6.11 Final Entity.80

Chapter 7 Conclusion and Future Scope

You might also like