Scribd
Upload
4 views3 pages

Neural Networks

Machine Learning encompasses various algorithms, including Deep Learning, which utilizes neural networks with multiple layers. Key deep learning models include ANN for general purposes, CNN for image tasks, and R-CNN for object detection. Neural networks consist of interconnected layers that process data, employing activation functions to learn complex patterns and are used in applications such as predictive analytics and image recognition.

Uploaded by

THIRUNEELAKANDAN
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
4 views3 pages

Neural Networks

Machine Learning encompasses various algorithms, including Deep Learning, which utilizes neural networks with multiple layers. Key deep learning models include ANN for general purposes, CNN for image tasks, and R-CNN for object detection. Neural networks consist of interconnected layers that process data, employing activation functions to learn complex patterns and are used in applications such as predictive analytics and image recognition.

Uploaded by

THIRUNEELAKANDAN
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Machine Learning is a broader field that includes algorithms like decision trees, SVM, and

linear regression.
Deep Learning is a specialized area within ML that uses neural networks with many layers
(i.e., deep structures).
ANN, CNN, R-CNN are all deep learning models:
 ANN – General-purpose deep learning model.
 CNN – Specialized for image-related tasks.
 R-CNN – Advanced CNN for object detection.

Neural Networks
A neural network is a computational model inspired by the way the human brain works. It is
made up of layers of interconnected nodes (neurons) that process and learn from data.
Basic Components:
1. Input Layer – Takes the raw data.
2. Hidden Layers – Perform calculations, extract patterns.
3. Output Layer – Gives the final prediction or classification.
Each neuron receives inputs, applies weights, adds bias, and passes the result through an
activation function (like ReLU or Sigmoid).
Activation Function
An activation function decides whether a neuron should be activated or not by transforming
the input signal into an output.
It adds non-linearity to the model, allowing the network to learn complex patterns.
Why It's Important
Without activation functions, a neural network would behave like a plain linear model—
unable to solve complex problems like image or voice recognition.
Used For:
 Predictive analytics
 Image and speech recognition
 Natural language processing
 Medical diagnostics
How It Learns:
 Forward Propagation – Data flows through the network to get a prediction.
 Loss Calculation – Measures the error in prediction.
 Backpropagation – Adjusts weights to reduce error using algorithms like gradient
descent.

ANN – General-purpose neural network.


CNN – Focused on image understanding.
R-CNN – Detects where objects are in images.
1. ANN (Artificial Neural Network)
 Concept: Mimics the way human neurons work.
 Structure:
o Input layer: Takes the data.
o Hidden layers: Perform computations using weighted connections.
o Output layer: Produces the final result.
 Learning: Uses backpropagation and gradient descent to minimize error.
 Applications:
o Stock price prediction
o Medical diagnosis
o Customer churn prediction

2. CNN (Convolutional Neural Network)


 Concept: Designed to recognize patterns in grid-like data such as images.
 Structure:
o Convolution layers: Apply filters to detect edges, shapes, textures.
o Pooling layers: Reduce spatial dimensions (e.g., max pooling).
o Fully connected layers: Perform classification based on extracted features.
 Advantages: Handles translation, scaling, and distortion in images well.
 Applications:
o Image and video classification
o Face recognition
o Medical image analysis

3. R-CNN (Region-based Convolutional Neural Network)


 Concept: Extends CNN for object detection, not just classification.
 Working:
1. Region Proposal: Uses selective search to find candidate object regions.
2. Feature Extraction: Applies CNN to each region.
3. Classification: Predicts object class and bounding box using SVM +
regression.
 Variants:
o Fast R-CNN
o Faster R-CNN
o Mask R-CNN (adds segmentation)
 Applications:
o Autonomous driving
o Surveillance
o Smart retail (product tracking)

 ANN = Think → answer.


 CNN = See → recognize.
 R-CNN = See → locate → recognize.

You might also like