netflow.js

netflow.js is a javascript library for visualizing neural network structure and training on the browser.

The library is ment to be used for explanation purposes.

You can see a demo here

Getting Started:

netflow.js is splited into multiple js files. You need to add the js files as shown below to your html file.

netflow.js depends on p5.js library for drawing on the html canvas.

<html>
  <head>
    <link rel="stylesheet" type="text/css" href="css/style.css">
    
    <script src="js/lib/p5.min.js"></script>
    <script src="js/values/style.js"></script>
    <script src="js/values/dimes.js"></script>
    <script src="js/utils/helper.js"></script>
    <script src="js/core/activation.js"></script>
    <script src="js/core/view.js"></script>
    <script src="js/core/draw.js"></script>
    <script src="js/core/netflow.js"></script>
    <script src="js/builder.js"></script>

  </head>
  <body>
    
  </body>
</html>

Building a Neural Network

To build a neural network, go to builder.js

Below is a sample 3-layer (input (2), hidden (2) and output (1) ) neural network. The first added layer is always the input layer, while the last is the output. Any layer added in between is a hidden layer.

function build(){
    var net = new Net();

    net.addLayer(2); // input 
    net.addLayer(2); // hidden
    net.addLayer(1); // output

    return net;

}

Viewing the Built Neural Network

Open index.html in the browser to view the built network.
To run the forward pass and backpropagation click on the button on the top left corner or press space bar.

Adjusting the Speed

To adjust the learning speed go to `dimes.js.
Change the value of DELAY variable.
The value is in milliseconds.

Adding Activation Function

You can add an activation function to a layer while building the network in builder.js file.