/NeuronNet

A neural networking library written in C#

Primary LanguageC#GNU Affero General Public License v3.0AGPL-3.0

What is NeuronNet ?

A fast, lightweight and easy to use neural network library written in C#.

Usage

//Create a new NetworkFactory
var networkFactory = new Network.Factory();

//Set input size to 2
networkFactory.InputSize = 2;

//Add 2 layers
networkFactory.AppendLayer(LayerType.Sigmoid, 2);
networkFactory.AppendLayer(LayerType.Sigmoid, 1);

//Build the network
var network = networkFactory.Build();

//Generate training data
var input = new double[] { 0.123, 0.3466 };
var output = new double[] { 0.723 };

while (true)
{
      //Train the network
      network.Train(input, output, 0.01);

      //Print out the error of the network
      Console.WriteLine("Error: " + network.GetError(input, output));
}

Or take a look at the sample project here.

How it works

The neural network is trained using backpropagation.

If you are interested, you can watch this youtube video (It's really awesome):

Backpropagation in 5 Minutes