Professional Documents
Culture Documents
to
Artificial Neural Systems – An Evolution
Email: vshrao@yahoo.com
URL: http://www.geocities.com/vshrao
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 1 / 30
FUNDAMENTAL OBJECTIVE OF COMPUTING IS
Intelligent System:
continually acts, mentally and externally, and by acting reaches its objectives
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 2 / 30
Nature’s Influence on Scientific Investigations
Nature is influencing researchers in many ways. History shows that many scientific
investigations were motivated by natural phenomena.
INSECTS ROBOTIC
MOVEMENTS
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 3 / 30
Nature’s Influence on Scientific Investigations (Contd…)
FISH SHIP
RESISTANCE
SPIDER MATERIALS
WEB
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 4 / 30
Bio-inspired Computing:
focuses on set of techniques inspired by biological sciences. Biological organisms often exhibit
properties that would be desirable in computer systems. Some features of bio-inspired
computing include:
tackling complex problems using computational methods modeled after design principles
encountered in nature.
producing informatics tools with enhanced robustness, scalability, flexibility, and which
can interface more effectively with humans.
The development and progress of computer science, engineering and technology has greatly
contributed to the study of biological systems and sciences to gain a better understanding of
biological processes and functions through the modeling and simulation of natural systems.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 5 / 30
Bio-Inspired Techniques:
The Present day researcher generates new ideas by taking advantage of the examples provided
by the nature.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 6 / 30
The Neuron:
Cells of nervous system or nerve cells are commonly called neurons. The purpose of the neuron
is to transmit messages through an electrochemical process throughout the nervous system.
These cells come in different shapes and sizes.
The neuron consists of CELL BODY / SOMA, the AXON, DENDRITES and the NEURONAL
MEMBRANE
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 7 / 30
The Neuron
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 8 / 30
Connections between two Biological Neurons
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 9 / 30
The Biological Brain
The brain is a complex multi-processing system. To
simplify thinking process, regular activities are
routinized so that they require less brain activity
and we do not have to attend to them consciously:
The structure of the human brain differs from that of other animals in several important
ways. These differences allow for many abilities over and above those of other animals, such
as advanced cognitive skills. Much of human brain structure is similar to that of other
mammals. The human brain also has a massive number of synaptic connections allowing for
a great deal of PARALLEL PROCESSING
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 10 / 30
Biological Neural Nets:
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 11 / 30
Functioning of the Biological Neuron
The brain is a collection of about 10 billion interconnected neurons. Each neuron is a cell that
uses biochemical reactions to receive, process and transmit information.
The strengths of all received charges are added together through a process of spatial and
temporal summation.
Spatial summation occurs when several weak signals are converted in to a single large
one, while temporal summation converts a rapid series of weak pulses from one source in
to one large signal.
The aggregate input is then passed to the soma or cell body. The soma and enclosed
nucleus don’t play a significant role in the processing of incoming and outgoing data.
Their primary function is to perform the continuous maintenance required to keep the
neuron functional.
The part of Soma that does concern itself with the signal is Axon hillock. If the aggregate
input is greater than the axon hillock’s threshold value, then the neuron fires, and an
output signal is transmitted down the axon.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 12 / 30
Functioning of the Biological Neuron (Contd..)
The strength of the output is constant; regard less of whether the input was just above the
threshold or a hundred times as great. The output strength is unaffected by the many
divisions in the axon; it reaches each terminal button with the same intensity it had at the
axon hillock.
Each terminal button is connected to the other neurons across a small gap called a
Synapse. The physical and neuro-chemical characteristics of each synapse determine the
strength and polarity of the new input signal.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 13 / 30
A Schematic diagram of a biological neuron
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 14 / 30
The Brain and the Computer
BRAIN COMPUTER
The brain uses chemicals to transmit information. The computer uses electricity.
Memories in the brain grow by stronger synaptic Computer memory grows by adding computer chips.
connections.
It is much easier and faster for the brain to learnThe computer can do many complex tasks at the same
new things. time that are difficult for the brain.
The human brain has weighed in at about 3 pounds Computers have evolved much faster than the human
for about the last 100,000 years. brain. The technological developments have made
computers faster, smaller and more powerful.
There are no new or used parts for the brain. Some It is easy to update computer with new parts.
work is being done with transplantation of nerve
cells for certain neurological disorders such as
Parkinson’s disease.
There are diseases that affect the brain. Computer can get a ‘VIRUS’.
The brain is always changing and being modified. The computer only changes when a new
There is no ‘OFF’ for the brain. Even when an hardware/software is added or some thing is saved in
animal is sleeping, its brain is still active and the memory. There is an ‘OFF’ for a computer when
working. the power is turned off.
The brain is better at interpreting the outside The computer is faster at doing logical things and
world and coming up with new ideas. computations.
‘There is more we do not know about the brain, than what we know’
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 15 / 30
Artificial Neural Networks (ANNs)
are inspired by neuro - scientific studies of the structure and function of human brain
are adaptive systems that changes their structure based on external or internal
information that flows through the network.
are similar to the biological neural networks in the sense that functions are performed
collectively and in parallel by the units.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 16 / 30
Artificial Neural Networks (Contd…)
Thus ANN is
Inputs Outputs
Inputs Outputs
NN
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 17 / 30
Why use Neural Networks?
Neural networks, with their remarkable ability to derive meaning from complicated or
imprecise data, can be used to extract patterns and detect trends that are too complex to be
noticed by either humans or other computer techniques. A trained neural network can be
thought of as an "expert" in the category of information it has been given to analyze. This
expert can then be used to provide projections given new situations of interest and answer
"what if" questions.
Other Advantages
Adaptive learning: An ability to learn how to do tasks based on the data given for training
or initial experience.
Real Time Operation: ANN computations may be carried out in parallel, and special
hardware devices are being designed and manufactured which take advantage of this
capability.
Fault Tolerance via Redundant Information Coding: Partial destruction of a network leads
to the corresponding degradation of performance. However, some network capabilities may
be retained even with major network damage.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 18 / 30
Mathematical Modeling
System is the connected units or parts that form a whole and operate together.
Function of the system depends not only on the functions of the constituents but also on how
they connect to one another.
An open system is a system that is influenced by its environment, where as a closed system is
one isolated from its environment
The assumptions of a closed system under certain restrictions is a means to simplify the
analysis of system behavior.
System Theory is the study of system behavior. The behavior refers to external input /
output characteristics and internal state changes. The behavior can be described in terms of
qualitative or quantitative relationships among variables defined for the system. A system can
be imitated or simulated by something which we call a MODEL.
A system may be represented by a block diagram. In this representation, the system is treated
as a ‘black box’ whose internal behavior is opaque, and its behavior is described by external
input / output relationships.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 19 / 30
Mathematical Modeling in General
In scientific investigations, a model is constructed to explain a hypothesis or to simulate a
practical system (as shown in the fig. below)
Abstraction
Simulation
An experiment is performed to gather the relevant data for estimating model parameters or
evaluating the model. A model needs to be refined or replaced if it does not provide sufficiently
accurate results.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 20 / 30
Steps involved in Mathematical Modeling
Define the problem for which a model is developed.
Postulate the structure of the model, for example, in terms of block diagram.
Develop model equations and estimate model parameters from relevant data.
Input variables (exogenous variables) which affect the system but whose behavior is not modeled.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 21 / 30
Neural Networks and Mathematical Modeling
Given a physical system, a neural network can model it on the basis of a set of examples
encoding the input / output behavior of the system.
The neural network learns to map an input into a desired output by self-adaptation.
The modeling capability of the neural network can be ascribed to its ability to learn the
mathematical function underlying the system operation.
a preprocessor
a postprocessor
a mathematical model
a baseline control
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 22 / 30
Integration of neural networks with other modeling techniques
A common practice is to select a tentative set of variables and refine the set successively.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 23 / 30
Correspondence between neural networks and mathematical models
………
Hidden layer
….…….
Input layer
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 24 / 30
Applications of ANNs:
Neural nets have broad applicability to real world problems. Some areas include
Classification
Noise Reduction
Sales forecasting
Industrial process control
Customer research
Data validation
Risk management
Target marketing
Pattern recognition
Diagnosis of hepatitis
Undersea mine detection
Texture analysis
3-D object recognition
Hand-written word recognition
Estimation of the rate of spread of an epidemic
Facial recognition
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 25 / 30
Are there any limits to Neural Networks?
The major issues of concern today are the scalability problem, testing, verification, and
integration of neural network systems into the modern environment.
Neural network programs sometimes become unstable when applied to larger problems. The
defense, nuclear and space industries are concerned about the issue of testing and verification.
The mathematical theories used to guarantee the performance of an applied neural network
are still under development. The solution for the time being may be to train and test these
intelligent systems much as we do for humans.
instability to explain any results that they obtain. Networks function as "black boxes”
whose rules of operation are completely unknown.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 26 / 30
The Future
Because gazing into the future is somewhat like gazing into a crystal ball, so it is better to
quote some "predictions". Each prediction rests on some sort of evidence or established trend
which, with extrapolation, clearly takes us into a new realm.
Prediction 1:
Neural Networks will fascinate user-specific systems for education, information processing,
and entertainment. "Alternative realities", produced by comprehensive environments, are
attractive in terms of their potential for systems control, education, and entertainment. This is
not just a far-out research trend, but is something which is becoming an increasing part of our
daily existence, as witnessed by the growing interest in comprehensive "entertainment centers"
in each home.
This "programming" would require feedback from the user in order to be effective but simple
and "passive" sensors (e.g. fingertip sensors, gloves, or wristbands to sense pulse, blood
pressure, skin ionization, and so on), could provide effective feedback into a neural control
system. This could be achieved, for example, with sensors that would detect pulse, blood
pressure, skin ionization, and other variables which the system could learn to correlate with a
person's response state.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 27 / 30
The Future (Contd...)
Prediction 2:
Neural networks, integrated with other artificial intelligence technologies, methods for direct
culture of nervous tissue, and other exotic technologies such as genetic engineering, will allow
us to develop radical and exotic life-forms whether man, machine, or hybrid.
Prediction 3:
Neural networks will allow us to explore new realms of human capability realms previously
available only with extensive training and personal discipline. So a specific state of consciously
induced neuro-physiologically observable awareness is necessary in order to facilitate a man
machine system interface.
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 28 / 30
John von Neumann
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 29 / 30
V. Sree Hari Rao, Jawaharlal Nehru Technological University, Hyderabad – 500 072 Page
Email: vshrao@yahoo.com; URL: http://www.geocities.com/vshrao 30 / 30