public static void Run(BrainWindow brainWindow)
{
var random = new System.Random(1234);
NetworkModel model = new NetworkModel(random);
model.AddLayer(NeuralLayer.Create(2, random, "INPUT"));
model.AddLayer(NeuralLayer.Create(1, random, "OUTPUT"));
brainWindow.Content = new NetworkModelView(model);
model.Build();
Console.WriteLine("----Before Training------------");
model.Print();
Console.WriteLine();
NeuralData input = new NeuralData(4);
input.Add(0, 0); //1
input.Add(0, 1); //2
input.Add(1, 0); //3
input.Add(1, 1); //4
NeuralData expectedOutput = new NeuralData(4);
expectedOutput.Add(0); //1
expectedOutput.Add(0); //2
expectedOutput.Add(0); //3
expectedOutput.Add(1); //4
// overall net learning rate
var eta = 0.15;
//alpha momentum, multiplier of last deltaWeight, [0.0..n]
var alpha = 0.5;
model.Train(input, expectedOutput, 100, eta, alpha);
Console.WriteLine();
Console.WriteLine("----After Training------------");
model.Print();
}
private void ShowBrainExecute()
{
if (_agent == null)
{
return;
}
var window = new BrainWindow {DataContext = new AgentBrainViewModel(_agent)};
window.ShowDialog();
}
private void ShowBestAgent(AgentType agentType)
{
if (SelectedAge == null)
{
MessageBox.Show("Age should be selected", "Information", MessageBoxButton.OK, MessageBoxImage.Information);
return;
}
var bestAgent = SelectedAge.GetLandscape().Statistics.GetMaxAgeAgent(agentType);
if (bestAgent == null)
{
MessageBox.Show("The best " + agentType + " is not determined yet", "Information", MessageBoxButton.OK, MessageBoxImage.Information);
return;
}
var viewModel = new AgentBrainViewModel(bestAgent);
var window = new BrainWindow();
window.DataContext = viewModel;
window.ShowDialog();
}
