public NeuralNetwork(int nInput, int nHidden, int nOutput)
{
vectorDisplayValue = new List <string>();
vectorDisplayValue.Add("Starte Initialisierung des neuronalen Netzes...");
this.nInput = nInput;
this.nHidden = nHidden;
this.nOutput = nOutput;
inputs = new double[nInput];
biasHidden = new double[nHidden];
outputs_h = new double[nHidden];
biasOut = new double[nOutput];
outputs = new double[nOutput];
W_IH = NeuralMath.MakeMatrix(nInput, nHidden);
W_HO = NeuralMath.MakeMatrix(nHidden, nOutput);
}
public double[] Feedforward(double[] xValues)
{
double[] hSums = new double[nHidden];
double[] oSums = new double[nOutput];
for (int i = 0; i < xValues.Length; ++i)
{
inputs[i] = xValues[i];
}
for (int j = 0; j < nHidden; ++j)
{
for (int i = 0; i < nInput; ++i)
{
hSums[j] += inputs[i] * W_IH[i][j];
}
}
for (int i = 0; i < nHidden; ++i)
{
hSums[i] += biasHidden[i];
}
for (int i = 0; i < nHidden; ++i)
{
outputs_h[i] = NeuralMath.HyperTan(hSums[i]);
}
vectorDisplayValue.Add("Werte im Hidden-Layer:");
List <string> hOutputsList = vectorDisplay.ShowVector(outputs_h, 4);
foreach (string element in hOutputsList)
{
vectorDisplayValue.Add(element);
}
for (int j = 0; j < nOutput; ++j)
{
for (int i = 0; i < nHidden; ++i)
{
oSums[j] += outputs_h[i] * W_HO[i][j];
}
}
for (int i = 0; i < nOutput; ++i)
{
oSums[i] += biasOut[i];
}
double[] softOut = NeuralMath.Softmax(oSums);
for (int i = 0; i < outputs.Length; ++i)
{
outputs[i] = softOut[i];
}
double[] result = new double[nOutput];
for (int i = 0; i < outputs.Length; ++i)
{
result[i] = outputs[i];
}
return(result);
}