0% found this document useful (0 votes)
28 views14 pages

Understanding Artificial Neural Networks

Assignment

Uploaded by

troyleo13
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
28 views14 pages

Understanding Artificial Neural Networks

Assignment

Uploaded by

troyleo13
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd

Go, change the

world

Name : Darshan C M Pawar


USN : RVCE21MIT010
Topic : Artificial Neural Networks
Go, change the
world

Agenda
• Basic Concept of ANN
• ANN Elements
• ANN Characteristics
• ANN Implementation
• ANN Architecture
• Applications
Go, change the
world

Basic Concept of ANN


● ANN is replica of neuron system in human brain. The human brain is
composed by billions neuron which are interconnected each others. A
biological neuron consists of three main components:
● Dendrites, that are input signals.
● Cell Body, where computation of input signals.
● Axon, is part which transmit output signals to another neuron.
Go, change the
world

Basic Concept of ANN


Go, change the
world

Basic Concept of ANN


● Dendrites are represented as input signal where data
such as feature or variables for problem solving will be
provided here.
● Cell Body is where linear combination computation
and activation function f() is applied.
● Axon is represented as output signal where result of
computation process in cell body will be delivered to
another neurons
Go, change the
world
ANN Elements
Based the explanation above, ANN element can be defined as:
• Input Layer, consists of neurons which represented features or variables
for problem solving.
• Output Layer, consists of neurons which represented result of
calculation in cell body
• Weights, is strength of connection between neurons
• Activation function, is function to obtain output according problem
which be solved
• Learning Function, or optimization function is to be used to obtain
minimum error by updating weights.
• Hidden layer, is optional depends on ANN architecture.
Go, change the
world
ANN Characteristics
● ANN can solve the problem which has pattern.
● ANN has good mapping ability, ANN can map the pattern on the input to
output which is connected continuously without interruption.
● ANN is a method that focuses on the learning process, ANN will be
trained using in-sample data to recognize the pattern. Based on result of
this train, ANN will be tested by using out sample data to identify the
pattern. Error of in-sample and out-sample data should be similar.
● ANN is tolerant of various types of data, it is able to identify patterns of
incomplete, partial or noisy data.
Go, change the
world

ANN Implementation

1. Regression, estimate relationship between independent


variable to dependent variable to obtain fitted function
according to data.
Go, change the
world

ANN Implementation
2. Classification, predict class of input data according to
target or data label.
Go, change the
world

ANN Implementation

3. Clustering, grouping set of data without data label or target


Go, change the
world
ANN Implementation
4. Forecasting, predict future condition using present and
past data or known as time series data.
Go, change the
world

ANN Architectures

ANN architecture or structure is a description of layer arrangement.


There are 3 types
● Single layer feedforward network
● Multilayer feedforward network
● Recurrent Network
Go, change the
world

Applications
Go, change the
world

Thank you!

You might also like