Please enable JavaScript.
Coggle requires JavaScript to display documents.
Neural Network (Units of neural network (Activation, Weight, Links,…
Neural Network
Units of neural network
Activation
Weight
Links
Activation function
Nodes(Units)
Bias weight
Why use neural networks?
Ability to derive meaning from complicated or imprecise data
Trained neural network can be thought as as expert
Adaptive learning
Self-organization
Real time operation
Fault Tolerance via Redundant Information Coding
Historical Backgrounds
Its simulation appear to be a recent development
The field survived a period of frustration and disrepute by following initial period of enthusiasm
First artificial neuron was introduced in 1943
Learning
Supervised Learning
Backpropagation
Requires a teacher that can calculate the desired output for given inputs
Back propagation networks are necessarily multi layer
perceptrons
Errors propagate backwards from output nodes to inner nodes
Unsupervised Learning
Hebbian Learning
Introduction
Information processing paradigm which is inspired from biological nervous system i.e. Brain
It is configured for specific application like pattern recognition or data classification
Network Structures
Feed-forward networks
Single-layer neural networks(perceptrons)
Multi-layer neural networks(perceptrons)
Feedback networks(Recurrent networks)