Scribd
Upload
13 views10 pages

Neural Network Presentation

Uploaded by

schaturvedi1603
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
13 views10 pages

Neural Network Presentation

Uploaded by

schaturvedi1603
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

Neuron and Nerve Structure

• Neuron:
• - Fundamental unit of the nervous system.
• - Consists of dendrites, axon, cell body (soma).

• Nerve Structure:
• - Bundle of axons.
• - Enclosed in connective tissues
(endoneurium, perineurium, epineurium).
Synapse
• Synapse:
• - Junction between two neurons or a neuron
and a target cell.

• Types:
• - Electrical Synapse
• - Chemical Synapse

• Process:
Artificial Neuron and Its Model
• Artificial Neuron:
• - Mathematical model inspired by biological
neurons.

• Model Components:
• - Input weights (w).
• - Bias (b).
• - Activation function (f).
• - Output (y = f(Σwx + b)).
Activation Functions
• Common Activation Functions:
• - Sigmoid
• - Tanh
• - ReLU (Rectified Linear Unit)
• - Leaky ReLU
• - Softmax
Neural Network Architecture
• Types:
• - Single Layer Neural Networks
• - Multi-Layer Neural Networks
• - Recurrent Neural Networks (RNNs)

• Components:
• - Input Layer
• - Hidden Layer(s)
• - Output Layer
Single and Multilayer Feed Forward
Networks
• Single-Layer Network:
• - One input layer and one output layer.

• Multilayer Network:
• - Includes one or more hidden layers.
• - More powerful for complex problems.
Recurrent Networks
• Recurrent Neural Networks (RNNs):
• - Connections form directed cycles.
• - Suitable for sequential data.

• Applications:
• - Time-series prediction
• - Natural language processing
Various Learning Techniques
• Types:
• - Supervised Learning
• - Unsupervised Learning
• - Reinforcement Learning
• - Semi-supervised Learning

• Approaches:
• - Gradient Descent
• - Backpropagation
Perceptron
• Perceptron:
• - Simplest type of artificial neural network.
• - Linear classifier.

• Algorithm:
• - Calculate weighted sum.
• - Apply activation function.
Convergence Rule
• Convergence Rule:
• - A learning rule ensuring weights stabilize
after repeated updates.
• - Perceptron Convergence Theorem:
• For linearly separable data, perceptron will
converge to a solution in finite steps.

You might also like