You are on page 1of 10

ARTIFICIAL NEURAL

NETWORKS
- SANTHOSH MAHENDRAN 9A
TABLE OF CONTENTS

01 02 03
Introduction To Neural How do Neural Networks
Network What Are Artificial Neural Work ?
Networks (ANN) ?

04 05
NNN vs ANN Structure Of ANN
Introduction To Neural
Networks
We can think of ANN as an artificial human brain system.
Neural Networks are of 2 types:
● BNN or NNN (Biological/Natural Neral System)
● ANN (Artificial Neural Network)
They can learn from their past experiences and are used in more
common everyday items than we think such as from face
recognition or text-to-speech to Self-Driving Cars.
What Are ANNs ?

● Artificial Neural Networks (ANNs) mimic the brain's structure using


digital neurons also called units. These units are arranged in layers
and connected, forming a network that processes information.
● They can learn from examples and also from experience. For
example, If you show them tons of cat pictures the next time you
show them a cat they will be easily able to recognize that it is a cat.
This is also why they are used in those application in which history
of experience with the input given by user is required such as facial
recognition, text-to-speech, language translation, image recognition
etc.
How Do ANNs Work ?
● Within an Artificial Neural Network (ANN), neurons
collaborate by forming intricate patterns of connections.
Neurons are like interconnected puzzle pieces, where the
output of one neuron serves as the input for others, creating a
dynamic network.
● The network's structure can be visualized as a directed,
weighted graph. Each connection between neurons has a
weight, representing the strength or importance of the
relationship. These weights are adjusted during training to
optimize the network's performance.
● Unlike traditional linear models, ANNs excel at capturing non-
linear relationships in data. The ability to process information
through multiple layers and varied connections enables ANNs
to model intricate relationships, making them well-suited for
tasks where complexities are involved.
Structure Of NNNs

Neurons Facilitation
These are the networks of Biological neural networks exhibit
interconnected neurons in living parallel processing, adaptability, and
organisms, particularly in the brain can learn from experiences through
and nervous system. mechanisms like synaptic plasticity.

Communication
Neurons in BNN communicate
through electrochemical signals,
forming a complex network that
processes information and controls
various functions.
Structure Of ANNs

Units Arrangement
These are the basic units that process and Neurons are organized into layers
transmit information. Neurons receive within the network. The input layer
inputs, perform a computation, and receives initial data, hidden layers
produce an output that becomes the input
process information, and the output
for other neurons in the network.
layer generates the final result.

Connection
Neurons are interconnected by
weighted links known as synapses.
These weights determine the strength
of connections and are adjusted
during training to facilitate the
learning process.
NNN vs ANN
NNN ANN
● ANN refers to artificial neural
● NNN refers to biological neural networks, which are the interconnected
networks, which are the artificial neurons called units in the
interconnected neurons in the brain system of applications and computers
and nervous system of living that are arranged in many layers.
organisms ● ANNs are programmed to be able to
● NNNs are naturally evolved learn from past experiences and
systems that inherently learn from recognize something already seen by it.
experiences and adapt to ● The have units which are arranged in
environmental stimuli. many layers such as the input layer,
● They have neurons which are middle layer and the output layer.
connected to each other and form ● The can give output in nanoseconds
nerves. range.
● They can respond in millisecond
range.
Functions Of ANNs
Four main functions of Artificial Neural Networks (ANNs):

1. ANNs excel at recognizing complex patterns in data, making them


invaluable for tasks like image and speech recognition.

2. ANNs are widely used for categorizing data into different classes
or groups based on learned features, making them powerful for
classification tasks.

3. ANNs can predict numerical values, making them suitable for


regression tasks such as forecasting stock prices or other quantitative
predictions.

4. ANNs contribute significantly to NLP applications, including


language translation, sentiment analysis, and speech understanding.
THANKS!

CREDITS: This presentation template was created by


Slidesgo, and includes icons by Flaticon, and infographics
& images by Freepik

You might also like