Neural networks introduction

Computational Intelligence provides us the opportunity to find asolution for the problems which were merely solvable by humanintelligence.
Computational intelligence machine can learn and remember similar tohuman brain Although the processor elements of a computer (semi-conductors) actmuch faster than processor elements of human brain (neurons),human response is faster than a computer.
In human brain, neurons work in parallel and are tightly connectedtogetherIn computer the calculations are doing sequential.
Artificial neural networks emulates brain capability of calculations anddecision making.
The simplest unit of neural networks named neurons Neurons transfer the information from sense organs to brain and frombrain to moving organs Each neuron is connected to other neurons and they totally make theneural network system.
There are more than 100 billion neurons in human body which mostof them are located in brain.
A biological brain includes three main parts: Dendrites: Receive signals from other neurons.
The neurotransmitter chemicals are released to transmitted the signalsthrough synaptic gaps Soma or body of the cell which accumulates all input signals.
When the input signals reach an action potential threshold, they aretransmitted to other neurons through Axon http : //people.eku.edu/ritchisong /301images/synapseN IAAA.gif Each neuron can adapt itself with environment changes The neural network structure is changing based on reinforcement andweakening the synaptic connections.
Learning is obtained by changing the synaptic gaps.
Artificial neural networks is inspired by biological neural networks.
So the structure of artificial neural networks are based on: Simple elements named neurons where information is processed.
Signals are transformed through the connections between neurons.
To each connection a weight is assigned which is multiplied to thetransferring signal.
At each neuron there is an activation function which is normally anonlinear function. This function provides the output of the neuron.
Each artificial neural network (NN) is distinguished by Pattern of connection between neurons (Neural network structure)Method of defining weights (Learning)Activation function By adjusting the weights, ( synaptic gaps in biological neurons) theneural network learn a pattern.
How much the artificial neural networks are similar to the biologicalneural networks? It varies in different type of artificial neural networks based on itsapplication.
For some researchers such as engineers high performance of the networkin calculations and function approximation is more important.
In some research areas like neurology, emulating the biological behavioris more attractive.
In general the artificial NNs and biological neural networks are similar in 1. The processing elements (neurons) receive signals2. Signals can be modified by weights (synaptic gaps)3. Processing elements gather the weighed inputs4. Under specified condition, the neuron provides output signal5. Output of a neuron can be transferred to other neurons6. The power of each synapse (weights) varies in different experience.
LearningParallel ProcessingGeneralization When a NN is trained, it can generalized its knowledge to the inputswhich has not seen beforeFor example if a NN is used for recognizing letters, if it receive a noisyinput, it still can recognize it and deliver the letter without noise.
NN can tolerate its malfunctioning in some circumstances.
Human is born with 100 billion neurons which some of them die butlearning does not stop!!Artificial NN should behave the same.
Such as eliminating echo on telephone lines 2. Control (NN can be applied for nonlinear systems) Identification, unmodeled dynamics, variable parametersObservationControl of nonlinear system Help in diagnosing diseases based on symptoms In classic methods, some rules are defined for standard pronunciation ofletters and a look-up table for exceptions.
In NN, there is no need to extract the rules and exceptions. NN istrained based on I/o data.
Introduction to Artificial Neural Systems, J. K. Zurada, West publishingcompany, 2nd edition 2006 Neural networks and learning machines, S. S. Haykin, Prentice Hall ,third edition,2008Fundamentals of Neural Networks, M. B. Menhaj, Amirkabir Universityof Technology, 2009 (in Farsi) concepts and models of NN)Single Layer Perceptron, Feed-forward NetworksRadial Bases Functions

Source: http://ele.aut.ac.ir/~abdollahi/Lec_1_NN.pdf