neural networks1
TRANSCRIPT
![Page 1: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/1.jpg)
Introduction to Neural Networks
![Page 2: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/2.jpg)
DEFINITION OF NEURAL NETWORKS
According to the DARPA Neural Network Study
(1988, AFCEA International Press, p. 60):
• ... a neural network is a system composed of
many simple processing elements operating in
parallel whose function is determined by network
structure, connection strengths, and the
processing performed at computing elements or
nodes.
![Page 3: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/3.jpg)
According to Haykin (1994), p. 2:
• A neural network is a massively parallel
distributed processor that has a natural
propensity for storing experiential knowledge
and making it available for use. It resembles
the brain in two respects:
– Knowledge is acquired by the network through a learning process.
– Interneuron connection strengths known as synaptic weights are used to store the knowledge
![Page 4: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/4.jpg)
BRAIN COMPUTATIONThe human brain contains about 10 billion nerve cells, or neurons. On average, each neuron is connected to other neurons through about 10 000 synapses.
![Page 5: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/5.jpg)
INTERCONNECTIONS IN BRAIN
![Page 6: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/6.jpg)
BIOLOGICAL (MOTOR) NEURON
![Page 7: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/7.jpg)
ASSOCIATION OF BIOLOGICAL NET WITH ARTIFICIAL NET
![Page 8: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/8.jpg)
The neuron is the basic information processing unit of a NN. It consists of:
1 A set of links, describing the neuron inputs, with weights W1, W2, …, Wm
2. An adder function (linear combiner) for computing the weighted sum of the inputs (real numbers):
3 Activation function : for limiting the amplitude of the neuron output.
m
1jj xwu
j
) (u y b
PROCESSING OF AN ARTIFICIAL NET
![Page 9: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/9.jpg)
BIAS OF AN ARTIFICIAL NEURONThe bias value is added to the weighted sum
∑wixi so that we can transform it from the origin.
Yin = ∑wixi + b, here b is the bias
x1-x2=0
x1-x2= 1
x1
x2
x1-x2= -1
![Page 10: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/10.jpg)
OPERATION OF A NEURAL NET
-
f
weighted sum
Inputvector x
output y
Activationfunction
weightvector w
w0jw1j
wnj
x0x1
xn
Bias
![Page 11: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/11.jpg)
![Page 12: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/12.jpg)
LAYER PROPERTIES•Input Layer:each input unit may be designated by an attribute value possessed by the instance
•Hidden Layer: not directly observable , provide nonlinearities for the network
•Output layer:encode possible values
![Page 13: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/13.jpg)
TRAINING PROCESSSupervised Training - Providing the network with a series of sample inputs and comparing the output with the expected responses
Unsupervised Training- Most similar input vector is assigned to the same output unit
Reinforcement Training- Right answer is not provided but indication of whether ‘right’ or ‘wrong’ is provided
![Page 14: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/14.jpg)
ACTIVATION FUNCTIONACTIVATION LEVEL – DISCRETE OR
CONTINUOUS
HARD LIMIT FUCNTION (DISCRETE) Binary Activation functionBipolar activation functionIdentity function
SIGMOIDAL ACTIVATION FUNCTION (CONTINUOUS)Binary Sigmoidal activation functionBipolar Sigmoidal activation function
![Page 15: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/15.jpg)
CONSTRUCTING ANN•Determine the network properties:
•Network topology•Types of connectivity•Order of connections•Weight range
•Determine the node properties:•Activation range
•Determine the system dynamics•Weight initialization scheme•Activation – calculating formula•Learning rule
![Page 16: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/16.jpg)
NEURAL NETWORKS
Neural Network learns by adjusting the weights so as to be able to correctly classify the training data and hence, after testing phase, to classify unknown data.
Neural Network needs long time for training.
Neural Network has a high tolerance to noisy and incomplete data
![Page 17: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/17.jpg)
SALIENT FEATURES OF ANN
Adaptive learning Self-Organisation Real Time Operation Fault Tolerance via Redundant
Information Coding Massive parallelism Learning and Generalizing Ability Distributed representation
![Page 18: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/18.jpg)
APPLICATIONS OF ANN
![Page 19: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/19.jpg)
REFERENCES1. P.D.Wasserman “Neural computing theory and
practice“ Van Nostrand Reinbold, 1989.
2. Welstead.S.T, “ Neural Networks and Fuzzy Logic Applications in C/C++”, John Wiley, New York, 1994
3. Yegnanarayana.B, “Artificial Neural Networks”, Prentice Hall of India Private Ltd, New Delhi, 1999
4. Zurada.J.M, “ Introduction to Aritificial Neural Systems”, Info Access and Distribution, Singapore, 1992
5. S.N.Sivanandam, S.Sumathi, S.N.Deepa, “Introduction to Neural Networks using MATLAB 6.0”, TMH Publishers, 2006.
![Page 20: Neural networks1](https://reader034.vdocuments.site/reader034/viewer/2022042723/5884bdb31a28ab34778b5ff5/html5/thumbnails/20.jpg)