Scribd
Upload
33 views10 pages

Unit 3 DLT

Uploaded by

sarmiladevin
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
33 views10 pages

Unit 3 DLT

Uploaded by

sarmiladevin
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
You are on page 1/ 10

Artificial Neural

Networks: A
Comprehensive
Guide
Artificial Neural Networks (ANNs) are powerful computational
models inspired by the structure and function of the human
brain. They consist of interconnected units called neurons,
organized in layers that process and learn from data. ANNs have
revolutionized various fields, including image recognition, natural
language
SH
processing, and machine learning.
Structure of Artificial Neural Networks
Input Layer Hidden Layers Output Layer

The input layer receives data Hidden layers transform the The output layer provides the
from the external world, which input data into a representation final output of the ANN,
the neural network needs to that is valuable for the output representing the network's
analyze or learn about. This data layer. The number of hidden response to the input data.
is then passed through hidden layers and units within each
layers. layer depends on the complexity
of the task.
Artificial Neurons vs
Biological Neurons
Feature Biological Neuron Artificial Neuron

Structure Cell body (soma), Input nodes,


dendrites, axon hidden layer
nodes, output
layer nodes
Synapses Connections Weights
between neurons connecting nodes

Learning Synaptic plasticity Backpropagation


algorithm

Activation Firing rate Activation function


How Artificial Neural
Networks Learn
ANNs are trained using a training set, which consists of labeled
examples. For instance, to teach an ANN to recognize a cat, it
is shown thousands of images of cats. The network learns to
identify patterns and features associated with cats by
adjusting the weights between neurons.
Types of Artificial Neural Networks

1 Feedforward Neural 2 Convolutional Neural 3 Modular Neural Network


Network Network (CNN)
A collection of independent
Data flows in a single CNNs use convolutional neural networks, each
direction, from the input layers to extract features performing a specific sub-
layer to the output layer, from images or other grid- task, working together to
with optional hidden layers. like data, making them achieve a complex output.
particularly useful for
computer vision tasks.

4 Radial Basis Function Neural Network (RBFN)5 Recurrent Neural Network (RNN)
RBFNs use radial basis functions to map input RNNs have feedback loops that allow them to
data to a higher-dimensional space, often used process sequential data, such as text or time
for modeling data with underlying trends. series, by remembering past information.
Applications of Artificial Neural Networks
Social Media Marketing and Sales Healthcare
ANNs power features like "People Personalized recommendations on ANNs are used in medical
you may know" on Facebook, e-commerce sites like Amazon and diagnosis, drug discovery, and
suggesting potential friends based Flipkart are driven by ANNs that personalized medicine, leveraging
on your profile and connections. analyze your browsing history and their ability to analyze complex
Facial recognition is another preferences. medical data and identify
application, using CNNs to identify patterns.
faces in images.
Linear Associative Networks
Linear associative networks, also known as linear associators,
are a fundamental concept in artificial neural networks. They
are known for their simplicity and ability to model basic
associative memory. Linear associative networks have a simple
two-layer structure, consisting of an input layer, a weight
matrix, and an output layer.
The Perceptron
A single-layer perceptron is the basic unit of a neural network.
It consists of input values, weights and a bias, a weighted sum,
and an activation function. The perceptron's output is
determined by the weighted sum of its inputs, which is then
passed through an activation function.
The Backpropagation Algorithm
Forward Pass
Input data is fed into the input layer and propagated
through the network, calculating activations at each
layer.

Error Calculation
The network's output is compared to the desired output
to calculate the error.

Backward Pass
The error is propagated backward through the network,
adjusting the weights of connections based on their
contribution to the error.
Deep Backprop Networks
Deep backpropagation networks, also known as deep neural
networks (DNNs), are powerful artificial neural network
architectures that utilize the backpropagation algorithm for
training. They achieve exceptional performance in various
tasks due to their ability to learn complex patterns from data.

You might also like