public void Initialize(
ActivationFunctions hiddenLayerActFunc = ActivationFunctions.Sigmoid,
ActivationFunctions outPutLayerActFunc = ActivationFunctions.Sigmoid)
{
//assign Layer dependencies
InputLayer.ChildLayer = HiddenLayer;
HiddenLayer.ParentLayer = InputLayer;
HiddenLayer.ChildLayer = OutputLayer;
OutputLayer.ParentLayer = HiddenLayer;
//set Activation Functions
HiddenLayer.ActivationFunction = hiddenLayerActFunc;
OutputLayer.ActivationFunction = outPutLayerActFunc;
//Assign random values to the Weights and set the bias
for (int i = 0; i < NrInputNeurons; i++)
{
InputLayer.NeuronBiasses[i] = 1;
InputLayer.BiasWeights[i] = Functions.RandomBias();
for (int j = 0; j < InputLayer.ChildLayer.NeuronValues.Length; j++)
{
InputLayer.SynapseWeights[i][j] = Functions.RandomBias();
}
}
for (int i = 0; i < NrHiddenNeurons; i++)
{
HiddenLayer.NeuronBiasses[i] = 1;
HiddenLayer.BiasWeights[i] = Functions.RandomBias();
for (int j = 0; j < HiddenLayer.ChildLayer.NeuronValues.Length; j++)
{
HiddenLayer.SynapseWeights[i][j] = Functions.RandomBias();
}
}
}
public Matrix ApplyDerActivation(Matrix input)
{
if (activationFunc == ActivationType.SIGMOID)
{
return(ActivationFunctions.DerSigmoid(input));
}
else if (activationFunc == ActivationType.TANH)
{
return(ActivationFunctions.DerTanh(input));
}
return(null);
}
// создание матриц весов и входных и выходных сигналов
void Create()
{
if (structure.inputs < 1)
{
throw new Exception("Create NeuralNetwork: inputs must be greater than zero");
}
if (structure.hiddens.Length == 0)
{
throw new Exception("Create NeuralNetwork: hiddens is null or zero");
}
for (int i = 0; i < structure.hiddens.Length; i++)
{
if (structure.hiddens[i] < 1)
{
throw new Exception("Create NeuralNetwork: hiddens at " + i + " layer must be greater than zero");
}
}
if (structure.outputs < 1)
{
throw new Exception("Create NeuralNetwork: outputs must be greater than zero");
}
layers = new Matrix[1 + structure.hiddens.Length];
inputs = new Vector[1 + structure.hiddens.Length];
outputs = new Vector[1 + structure.hiddens.Length];
layers[0] = new Matrix(structure.hiddens[0], structure.inputs);
for (int i = 0; i < structure.hiddens.Length - 1; i++)
{
layers[i + 1] = new Matrix(structure.hiddens[i + 1], structure.hiddens[i]);
}
layers[layers.Length - 1] = new Matrix(structure.outputs, structure.hiddens[structure.hiddens.Length - 1]);
hiddensActivation = ActivationFunctions.GetFunction(structure.hiddensFunction);
hiddensDerivative = ActivationFunctions.GetDerivative(structure.hiddensFunction);
outputActivation = ActivationFunctions.GetFunction(structure.outputFunction);
outputDerivative = ActivationFunctions.GetDerivative(structure.outputFunction);
}