public static void Main(string[] args)
{
nn = new NeuralNetwork(new int[] { 2, 4, 1 });
trainingData = new SupervisedData[4];
trainingData[0] = new SupervisedData(new Matrix(new float[, ] {
{ 0 }, { 0 }
}), new Matrix(new float[, ] {
{ 0 }
}));
trainingData[1] = new SupervisedData(new Matrix(new float[, ] {
{ 0 }, { 1 }
}), new Matrix(new float[, ] {
{ 1 }
}));
trainingData[2] = new SupervisedData(new Matrix(new float[, ] {
{ 1 }, { 0 }
}), new Matrix(new float[, ] {
{ 1 }
}));
trainingData[3] = new SupervisedData(new Matrix(new float[, ] {
{ 1 }, { 1 }
}), new Matrix(new float[, ] {
{ 0 }
}));
nn.Train(trainingData, 100000);
Console.WriteLine("Input = 0, 0");
Console.WriteLine("Output = " + nn.FeedForward(trainingData[0].GetInput()));
Console.WriteLine("Input = 0, 1");
Console.WriteLine("Output = " + nn.FeedForward(trainingData[1].GetInput()));
Console.WriteLine("Input = 1, 0");
Console.WriteLine("Output = " + nn.FeedForward(trainingData[2].GetInput()));
Console.WriteLine("Input = 1, 1");
Console.WriteLine("Output = " + nn.FeedForward(trainingData[3].GetInput()));
}
public static void Main(string[] args)
{
nn = new NeuralNetwork(new int[] { 3, 4, 5, 4, 1 }, 0.2f);
nn = NeuralNetwork.Load("Test");
trainingData = new ISupervisedData[8];
//Output = (A or B) and c
trainingData[0] = new SupervisedData(new Matrix(new float[, ] {
{ 0 }, { 0 }, { 0 }
}), new Matrix(new float[, ] {
{ 0 }
}));
trainingData[1] = new SupervisedData(new Matrix(new float[, ] {
{ 0 }, { 0 }, { 1 }
}), new Matrix(new float[, ] {
{ 0 }
}));
trainingData[2] = new SupervisedData(new Matrix(new float[, ] {
{ 0 }, { 1 }, { 0 }
}), new Matrix(new float[, ] {
{ 0 }
}));
trainingData[3] = new SupervisedData(new Matrix(new float[, ] {
{ 0 }, { 1 }, { 1 }
}), new Matrix(new float[, ] {
{ 1 }
}));
trainingData[4] = new SupervisedData(new Matrix(new float[, ] {
{ 1 }, { 0 }, { 0 }
}), new Matrix(new float[, ] {
{ 0 }
}));
trainingData[5] = new SupervisedData(new Matrix(new float[, ] {
{ 1 }, { 0 }, { 1 }
}), new Matrix(new float[, ] {
{ 1 }
}));
trainingData[6] = new SupervisedData(new Matrix(new float[, ] {
{ 1 }, { 1 }, { 0 }
}), new Matrix(new float[, ] {
{ 0 }
}));
trainingData[7] = new SupervisedData(new Matrix(new float[, ] {
{ 1 }, { 1 }, { 1 }
}), new Matrix(new float[, ] {
{ 1 }
}));
nn.Train(trainingData, 100000);
nn.Save("Test");
for (int i = 0; i < trainingData.Length; i++)
{
Console.Write("Input = \n" + trainingData[i].GetInput());
Console.Write("\nExpected = \n" + trainingData[i].GetOutput());
Matrix output = nn.FeedForward(trainingData[i].GetInput());
Console.Write("\nActual = \n" + output);
Console.WriteLine("\n");
}
}
