Novabrain

Novabrain is a javascript neural network library for Node.js and browser.
This library implements a multilayer perceptron network that you can train to learn XOR, OR, AND ... for example.

!Perceptron

#### In Node.js

You can install Novabrain with npm

``
$ npm install novabrain --save
`
`javascript
var Novabrain = require('novabrain');
var Neuron = Novabrain.Neuron;
var Layer = Novabrain.Layer;
var Network = Novabrain.Network;
var Trainer = Novabrain.Trainer;
var Transfer = Novabrain.Transfer;
var Samples = Novabrain.Samples;
`

#### In the browser

You can also use the minified version to increase your web page loading

`html

`

`html

`

$3

Constructor expected an intergers suite.
The first value is the input size
The last value is the output size
Between this values you can set many hidden size (2, 3, ..., 1)

`javascript
new Novabrain.Network(2,1);
new Novabrain.Network(2,3,1);
new Novabrain.Network(5,4,4,2);
`

$3

Novabrain samples contains training and config for basics functions

`javascript
Novabrain.Samples.XOR
Novabrain.Samples.AND
Novabrain.Samples.OR
`

$3

This example shows how the neural network is trained to learn XOR

`javascript
var network = new Novabrain.Network(2,1);
var trainer = new Novabrain.Trainer(network);

trainer.train([
{ input: [0,0], output: [0] },
{ input: [0,1], output: [1] },
{ input: [1,0], output: [1] },
{ input: [1,1], output: [0] },
]);

console.log([0,0], network.output([0,0])); // [~0.05]
console.log([0,1], network.output([0,1])); // [~0.93]
console.log([1,0], network.output([1,0])); // [~0.93]
console.log([1,1], network.output([1,1])); // [~0.09]

network.transfer = Novabrain.Transfer.BOOLEAN;

console.log([0,0], network.output([0,0])); // [false]
console.log([0,1], network.output([0,1])); // [true]
console.log([1,0], network.output([1,0])); // [true]
console.log([1,1], network.output([1,1])); // [false]
`

$3

The transfer functions are used to change the value of the outputs.
By default, neurons uses a Logistic Sigmoid transfer.
You can change those properties the following way.

`javascript
network.transfer = Novabrain.Transfer.BOOLEAN;

console.log([0,0], network.output([0,0])); // [false]
console.log([0,1], network.output([0,1])); // [true]
console.log([1,0], network.output([1,0])); // [true]
console.log([1,1], network.output([1,1])); // [false]
`

LOGISTIC
Return logistic sigmoid values

HARDLIMIT
Return 0 or 1 values

BOOLEAN
Return boolean values like HARDLIMIT

IDENTIFY
Return sum values without transfer

TANH
Return values between -1 and 1

$3

`javascript
var n1 = new Novabrain.Network(2,1);
var n2 = new Novabrain.Network(2,1);

n2.import(n1);
// or
n2.import(n1.export());

var results = n2.output([...]));
`

$3

By default the transfer function used is LOGISITC but you can change this by two ways.
Define your custom transfer before the standalone function export or set the transfer param when you use the standalone function.

`javascript
var standalone = network.standalone();
var booleanResults = standalone([...], Novabrain.Transfer.BOOLEAN));
`

`javascript
var standalone = network.standalone(Novabrain.Transfer.BOOLEAN);
var booleanResults = standalone([...]));
var tanhResults = standalone([...], Novabrain.Transfer.TANH));
`

$3

`
$ npm test
`

`
$ make tests
`

`
$ npm install mocha -g
$ mocha
`

$3

Novabrain is an Open Source project started in France by François Mathey.
Anybody is welcome to contribute to the development of this project.

If you want to contribute feel free to send PR's, just make sure to run the make before submiting it.
This way you'll run all the test specs and build the web distribution files.

`
$ make
``

Thank you <3