Introduction to Neural Networks
What is a Neural Networks ?
Loosely Based on how the neurons in your brain work
They have been particularly good at solving really big really complex problems
- Image classification
- Natural Language Processing
- Cat Classification
One of the neurons in the artificial neural network, trained from still frames from unlabeled YouTube
videos, learned to detect cats.
GoogleBlog
- Input Layer
- Output Layer
- Hidden Layer
Circles Represent
Neuron or Output or Input
Lines Represent
Synapse
The Synapse has a really simple job They take a input value multiple with their weight and output the result to the next neuron
Neuron Takes The Sum of all the outputs from the Synapse and applies a Activation function
You can combining multiple such hidden layers to create a Deep Belief Network.
There are two types of preceptron based Artificial Neural Network.
- Convolution Neural Network.
- Recurring Neural Network
Thank you