THE COOPER UNION
ALBERT NERKEN SCHOOL OF ENGINEERING
AN EXPLORATION AND DEVELOPMENT OF CURRENT
ARTIFICIAL NEURAL NETWORK THEORY AND APPLICATIONS
WITH EMPHASIS ON ARTIFICIAL LIFE
by
David J. Cavuto
A thesis submitted in partial fulfillment
of the requirements for the degree of
Master of Engineering
May 6, 1997
THE COOPER UNION FOR THE ADVANCEMENT OF SCIENCE AND ART
THE COOPER UNION FOR THE ADVANCEMENT OF SCIENCE AND ART
ALBERT NERKEN SCHOOL OF ENGINEERING
This thesis was prepared under the direction of the Candidate’s Thesis
Advisor and has received approval. It was submitted to the Dean of the
School of Engineering and the full Faculty, and was approved as partial
fulfillment of the requirements for the degree of Master of Engineering.
_________________________________
Dean, School of Engineering - Date
______________________________
Prof. Simon Ben-Avi - Date
Candidate’s Thesis Advisor
i
Acknowledgments
I would like to take this opportunity to thank, first and foremost, my thesis advisor, Dr. Simon
Ben-Avi. His advice, both as a professor and as a friend, were and always will be invaluable.
Moreover, I would like to thank the entire EE department faculty and staff for all the support
and encouragement (and toleration) they have shown me throughout the years.
I am deeply indebted to my friend and Big Brother Yashodhan Chandrahas Risbud (Yash!).
Without the occasional smack in the head he needed to give me, I might not have made it
through at all. Thanks for putting up with me.
Kappa Phi – Zeta Psi. My brothers supported me in the hard times and cheered me in the
good times. Can anyone ask for more?
Special thanks to The Leib, Seamous, and of course, my Muffin.
My utmost appreciation and thanks to my parents, George and Doris Cavuto. What can I
say? Thanks for everything. (Especially all that money!)
And finally, a big old THANKS! to Peter Cooper for giving me a place to work, learn, and
grow for the last six years. Anywhere else would have been just a school. The Cooper Union
has been my home.
– DJC
Disclaimer: this thesis is entirely a product of my imagination. Any resemblance to actual work is purely coincidental.
ii
1. Abstract
The purpose of this study is to explore the possibilities offered by current Artificial Neural Net
(ANN) structures and topologies and determine their strengths and weaknesses. The biological
inspiration behind ANN structure is reviewed, and compared and contrasted with existing
models. Traditional experiments are performed with these existing structures to verify theory
and investigate more possibilities. This study is conducted to the end of examining the
possibility of using ANNs to create “artificial life,” which is defined here as a structure or
algorithm which displays characteristics typically only attributed to biological organisms,
usually nonrepeating, nonrandom processes. Although some ANN topology is shown to be
highly similar to that of biological systems, existing ANN algorithms are determined be
insufficient to generate the desired type of behavior. A new ANN structure, termed a
“Temperon”, is designed, which encompasses more properties in common with biological
neurons than did its predecessors. A virtual environment based on turtle graphics is used as a
testbed for a neural net built with the new type of neuron. Experiments performed with the
Temperon seem to confirm its ability to learn in an unassisted fashion.
iii
Table of Contents
1. ABSTRACT ii
2. BACKGROUND 1
2.1 BIOLOGICAL NATURE OF NEURAL CELLS 1
2.1.1 PHYSICAL STRUCTURE OF BIOLOGICAL NEURON 1
2.1.1.1 Body, Axon, Dendrites, Synapse 1
2.1.1.2 Neurotransmitter 3
2.1.1.3 Sodium/Potassium Pump 4
2.1.1.4 Ionized pulse 6
2.1.1.5 All-or-Nothing Causation 8
2.1.2 MATHEMATICAL REPRESENTATION OF NERVE CELL PROCESSES 10
2.1.2.1 Mathematical correlation to the physical interconnections 10
2.1.2.2 Linear combination of inputs 11
2.1.2.3 Thresholding resulting in binary or near-binary outputs 12
2.2 ARTIFICIAL NEURAL NETS AND THEIR APPLICATIONS 14
2.2.1 GENERAL THEORY 14
2.2.1.1 Purpose 14
2.2.1.2 Structure 14
2.2.1.3 Weight Updating 15
2.2.2 PERCEPTRONS - CLASSIFICATION 15
2.2.2.1 Single Layer 15
2.2.2.2 MLP - Feedforward 18
2.2.3 HOPFIELD NET - PATTERN RECOGNITION 21
2.2.4 GENERALIZATIONS 22
2.3 OUR FRIEND APLYSIA 24
2.3.1 GENERAL OBSERVATIONS 24
2.3.2 SUMMARY OF RELEVANT EXPERIMENTS 25
2.3.2.1 Habituation 25
2.3.2.2 Sensitization 26
2.3.3 RELEVANCE AND RELATION TO NEURAL NETS 26
评论1