Matters of Discussion

Biological Inspiration
Artificial Neurons
and
Neural Networks Application
******************************************************************************************************************

Supervised vs. Unsupervised Learning:

Key concept and Differences

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan]

1
 Biological Neuron
Animals are able to react adaptively to changes in their external and internal
environment, and they use their nervous system to perform these
behaviours.
An appropriate model/simulation of the nervous system should be able to
produce similar responses and behaviours in artificial systems.

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan] 2

Biological Neuron
 Human brain contains 1010 basic units called neurons.
 Each neuron connects to 104 number of neurons.
 Neuron receives electrochemical signals from it’s
sources and responds by transmitting electrical
impulses to other neurons.
 Neuron weight = 1.5 ∗ 10−9 grams
 Part of neurons perform I/O operations.
 Part of neurons perform signal transformation and
information storage.
 However, all neurons have similar characteristics.

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan]

3
 Biological Neuron

 The information transmission happens at the synapses.

 Cell body is called as Soma or nucleolus.
 Dendrites as input channels [structure—branching tree]
 Link attached to soma is called axon that acts as output.
 Axon terminates with synaptic contacts called synaptic junction.

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan] 4

 Artificial Neurons
The fundamental building block of Machine Learning is the
Perceptron which is a single neuron in a Neural Network.

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan] 5

 Artificial Neural Networks[ANN/NN]
NN are the Simplified model of biological nervous
system.
Highly interconnected network of large number of
processing elements called neurons.
NN learns by example. So, it can be trained by known
example of a problem to acquire knowledge about it.
Once it is trained perfectly it can put for unknown
problem instances.
It adopts various learning mechanisms to acquire
knowledge from known examples.
1. Supervised learning mechanism
2. Unsupervised learning mechanism.
6
 Perceptron model

x1
x2 w1
x3 Output
n
w2
… z   wi xi ; y  H ( z )
Inputs

i 1
y
w3
xn-1 .
xn
.
.
An artificial neuron is
wn-1
inspired by biological
wn neurons known as
Perceptron.
one possible model

ARTIFICIAL NEURAL NETWORKS 7

 Neural Network Architecture
Single layer perceptron model
Multi-layer perceptron model

ARTIFICIAL NEURAL NETWORKS 8

 Single-layer Perceptron
Perceptron has just two layers of input layers and output
layers. Often called a single-layer network on account of
having 1 layer of links, between input and output.
Input nodes are connected fully to a node or multiple nodes
in the next layer. A node in the next layer takes a weighted
sum of all its inputs.

9
 Multi-layer perceptron model
Neural network: input / output transformation

X –inputs
W- weights
Output
Inputs

yout  F ( x,W )

An artificial neural network is composed of many artificial neurons that are linked
together according to a specific network architecture. The objective of the neural
network is to transform the inputs into meaningful outputs.

Artificial Neural Networks 10

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan]
11
Multi-Layer Perceptron (MLP)..Cont....
A multilayer perceptron (MLP) is a feed-forward
artificial neural network that generates a set of
outputs from a set of inputs.
An MLP is a neural network connecting multiple
layers in a directed graph, which means that the
signal path through the nodes only goes one way.
 The MLP network consists of input, output, and
hidden layers.
Each hidden layer consists of numerous
perceptron's which are called hidden layers or
hidden unit.

12
 Multi-Layer Perceptron (MLP)..Cont....

Input Layer Hidden Layers Output Layer

• Input Layer: - The Input Layers provide information from the

outside world (environment) to the network.
• Hidden Layers: - Hidden Layers perform computations and
transfer information from the input layer to the output layers.
Hidden layers have no direct connection with the outside
world.
• Output Nodes: – The Output Layer are responsible for
computations and transferring information from the network
to the outside function (Environment).
13
 Processing Information in ANN
Inputs Weights Outputs

x1
w1 Y1

x2 w2 Neuron (or PE) f (S )

. S  
n
X iW
Y
. Y2
. i 1
i

.
. Summation
Transfer
.
Function
wn Yn
xn

• A single neuron (processing element – PE) with

inputs and outputs
Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan]
14
 Tasks to be solved by artificial neural
networks:
 controlling the movements of a robot based on self-
perception and other information (e.g., visual
information);
 deciding the category of potential food items
(e.g., eatable or non-eatable) in an artificial
world;
 recognizing a visual object (e.g., a familiar face);
 predicting where a moving object goes, when a
robot wants to catch it.
Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan] 15
 ANN/NN TERMINOLOGIES, USE, APPS
 ANN/NN
 Neural computing / Neuro computers / connectionist
network / parallel and distributed processors.
Uses of ANN
 pattern recognition, forecasting, prediction, and
classification
Application areas of ANN
 finance, marketing, manufacturing, operations,
information systems, and so on

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan] 16

 Learning in ANN
• Learning is a NN training process of acquiring
knowledge.
• NN can acquire knowledge through learning.
• A process by which a neural network learns the
underlying relationship between input and outputs, or
just among the inputs
• Supervised Machine learning
– For classification type problems
– Assume an expert present during learning process
– relationship between input and outputs
• Unsupervised Machine learning
– For clustering type problems
– Self-organizing / no expert present
– Relationship just among the inputs are considered

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan]

17
 Cont..
In Supervised learning, you train the machine
using data which is well "labeled." It means
some data is already tagged with the correct
answer. It can be compared to learning which
takes place in the presence of a supervisor or
a teacher.
Unsupervised learning is a machine learning
technique, where you do not need to
supervise the model. Instead, you need to
allow the model to work on its own to
discover information.
18
 A Supervised Learning Process
ANN
Model
Three-step process:
1. Compute temporary
Compute
output
outputs
2. Compare outputs with
desired targets
3. Adjust the weights and
Is desired
Adjust
weights
No
output repeat the process
achieved?

Yes

Stop
learning

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan] 19

 How Supervised Learning works?
For example, you want to train a machine to help you predict how long it will
take you to drive home from your workplace. Here, you start by creating a set
of labeled data. This data includes
Weather conditions
Time of the day
Holidays
All these details are your inputs. The output is the amount of time it took to
drive back home on that specific day.

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan] 20

 pcai.com
• Knowledge Based Systems, AI Languages,
Neural Networks, Machine Learning, Genetic
Algorithms, Evolutionary Software, Expert
Systems, Fuzzy Logic, Data Mining, Intelligent
Agents, Business Rules, Case-Based
Reasoning, Common Sense, Data
Visualization, Inferencing, Forecasting, Pattern
Matching, Speech, Rule-Based Systems, Text
Mining, Vision, Robotics.
Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan]
21
Activity for you
Try to understand the concept of ANN…..

Investigate the ANN architecture and Machine

Learning processes…..

Design a report for that…..

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan]

22
 Cheers For the Great Patience!
Query Please?

Compiled By: Dr. Nilamadhab Mishra [(PhD- CSIE) Taiwan]

23