public void Instantiate() {
int[] actualLayersSizes = new int[LayersSizes.Length + 2];
actualLayersSizes[0] = 300;
actualLayersSizes[actualLayersSizes.Length - 1] = 10;
for (int i = 0; i < LayersSizes.Length; i++)
actualLayersSizes[i + 1] = LayersSizes[i];
Instance = new Perceptron(actualLayersSizes);
Instance.Initialize();
}
public void Instantiate()
{
int[] actualLayersSizes = new int[LayersSizes.Length + 2];
actualLayersSizes[0] = 300;
actualLayersSizes[actualLayersSizes.Length - 1] = 10;
for (int i = 0; i < LayersSizes.Length; i++)
{
actualLayersSizes[i + 1] = LayersSizes[i];
}
Instance = new Perceptron(actualLayersSizes);
Instance.Initialize();
}
static void Run(string[] args)
{
Engine.Output = new ConsoleStreamOutput();
if (args.Length == 0)
{
throw new Exception("error: invalid arguments\nUse /? or /help to see available commands");
}
INeuralNetwork net;
switch (args[0])
{
case "/?":
case "/help":
Console.WriteLine("This is help.");
break;
case "/data":
if (args.Length != 2)
{
InvalidArguments();
}
Settings.Instance.DataDirectoryPath = args[1];
Settings.Save();
break;
case "/tests":
if (args.Length != 2)
{
InvalidArguments();
}
Settings.Instance.TestsDirectoryPath = args[1];
Settings.Save();
break;
case "/rate":
if (args.Length != 2)
{
InvalidArguments();
}
Settings.Instance.LearningRate = double.Parse(args[1]);
Settings.Save();
break;
case "/create":
if (args.Length < 2)
{
InvalidArguments();
}
int[] layersSizes = new int[args.Length - 1];
for (int i = 0; i < layersSizes.Length; i++)
{
layersSizes[i] = int.Parse(args[i + 1]);
}
Engine.Create(layersSizes);
break;
case "/learn":
if (args.Length != 1)
{
InvalidArguments();
}
Engine.Learn();
break;
case "/classify":
if (args.Length != 2)
{
InvalidArguments();
}
Engine.Classify(args[1]);
break;
case "/optimize":
Genetics genetics = new Genetics();
net = genetics.Optimize();
net.Save(Settings.Instance.NetworkFileName);
Settings.Save();
break;
case "/inspect":
if (args.Length != 1)
{
InvalidArguments();
}
net = new Perceptron();
net.Load(Settings.Instance.NetworkFileName);
Console.WriteLine(net.Inspect());
break;
default:
InvalidArguments();
break;
}
}
static void Run(string[] args) {
Engine.Output = new ConsoleStreamOutput();
if (args.Length == 0)
throw new Exception("error: invalid arguments\nUse /? or /help to see available commands");
INeuralNetwork net;
switch (args[0]) {
case "/?":
case "/help":
Console.WriteLine("This is help.");
break;
case "/data":
if (args.Length != 2)
InvalidArguments();
Settings.Instance.DataDirectoryPath = args[1];
Settings.Save();
break;
case "/tests":
if (args.Length != 2)
InvalidArguments();
Settings.Instance.TestsDirectoryPath = args[1];
Settings.Save();
break;
case "/rate":
if (args.Length != 2)
InvalidArguments();
Settings.Instance.LearningRate = double.Parse(args[1]);
Settings.Save();
break;
case "/create":
if (args.Length < 2)
InvalidArguments();
int[] layersSizes = new int[args.Length - 1];
for (int i = 0; i < layersSizes.Length; i++)
layersSizes[i] = int.Parse(args[i + 1]);
Engine.Create(layersSizes);
break;
case "/learn":
if (args.Length != 1)
InvalidArguments();
Engine.Learn();
break;
case "/classify":
if (args.Length != 2)
InvalidArguments();
Engine.Classify(args[1]);
break;
case "/optimize":
Genetics genetics = new Genetics();
net = genetics.Optimize();
net.Save(Settings.Instance.NetworkFileName);
Settings.Save();
break;
case "/inspect":
if (args.Length != 1)
InvalidArguments();
net = new Perceptron();
net.Load(Settings.Instance.NetworkFileName);
Console.WriteLine(net.Inspect());
break;
default:
InvalidArguments();
break;
}
}