simpleneuron
v0.0.1
Published
Simple neural network library for browser and Node.js
Downloads
10
Readme
SimpleNeuron
Simple neural network library for browser and Node.js. Work in Progress.
Installation
Via npm on Node:
npm install simpleneuron
Usage
Reference in your program:
var sn = require('simpleneuron');
Create a neuron
var neuron = sn.neuron();
It is a simple threshold neuron with trigger value = 1.
Connect a neuron with other neuron
var neuron = sn.neuron();
var input = sn.neuron();
neuron.input(input);
The input has weight 1
Connect a neuron with other neuron using a weight value
var neuron = sn.neuron();
var input = sn.neuron();
neuron.input(input, -0.5);
The input has weight -0.5.
Set the output value of neuron:
neuron.output(1);
Usually, this method is used to set the initial neuron layer values.
Get the output value of a neuron:
var result = neuron.output();
Create a network of neurons, with three layers, containing 4, 15, 3 neurons. The weights are random values between -1 and 1:
var network = ss.network([4, 15, 3]);
Get the number of neurons in a network:
var count = network.neurons();
In the previous created network, the count is 4+15+3 == 22.
Get the number of neurons in a layer:
var count0 = network.neurons(0); // 4
var count1 = network.neurons(1); // 15
var count2 = network.neurons(2); // 3
TBD
Development
git clone git://github.com/ajlopez/SimpleNeuron.git
cd SimpleNeuron
npm install
npm test
Samples
TBD
Versions
- 0.0.1 Published
License
MIT
References
- Extreme Learning Machine
- Backpropagation
- Using neural nets to recognize handwritten digits
- How the backpropagation algorithm works
- Using neural nets to recognize handwritten digits
- THE MNIST DATABASE of handwritten digits
Books
- Neural Networks, Simon Haykin
- Fundamentals of Deep Learning, Nikhil Buduma (2017, O�Reilly)
Contribution
Feel free to file issues and submit pull requests � contributions are welcome<
If you submit a pull request, please be sure to add or update corresponding
test cases, and ensure that npm test
continues to pass.