public RBF() // Use it for Testing One Sample from The Saved Data
{
Class = new ReadData[4];
Class[0] = new ReadData("\\Dataset\\Training Dataset\\Closing Eyes", 1);
Class[1] = new ReadData("\\Dataset\\Training Dataset\\Looking Down", 2);
Class[2] = new ReadData("\\Dataset\\Training Dataset\\Looking Front", 3);
Class[3] = new ReadData("\\Dataset\\Training Dataset\\Looking Left", 4);
}
public MLP(double pEta, int pEpochs, bool pBias, List <int> pNumNodesperHiddenLayer, bool readfile)
{
open = false;
p = new Process();
numHiddenLayers = pNumNodesperHiddenLayer.Count;
accuracy = 0;
desiredOutputs_D = new int[4] {
0, 0, 0, 0
};
Errors = new double[4];
pNumNodesperHiddenLayer.Add(4); //num of weights_HO per output Layer
NumNodesperEachLayer_IHO = pNumNodesperHiddenLayer;
NumNodesperEachLayer_IHO.Insert(0, 19);
biasLst = new List <List <double> >();
NumOfWeightsPerEachLayer = new List <int>();
NetInputs = new List <List <double> >();
actualOutputs = new List <List <double> >();
Deltas = new List <List <double> >();
int numOfPreviousWeights = 19; //intial of numof weights_HO for input weights_HO
for (int i = 0; i < numHiddenLayers + 1; numOfPreviousWeights = pNumNodesperHiddenLayer[i + 1], i++)
{
NumOfWeightsPerEachLayer.Add((numOfPreviousWeights * pNumNodesperHiddenLayer[i + 1]));
}
Class = new ReadData[4];
Class[0] = new ReadData("\\Dataset\\Training Dataset\\Closing Eyes", 1);
Class[1] = new ReadData("\\Dataset\\Training Dataset\\Looking Down", 2);
Class[2] = new ReadData("\\Dataset\\Training Dataset\\Looking Front", 3);
Class[3] = new ReadData("\\Dataset\\Training Dataset\\Looking Left", 4);
Normalize_Data();
weights_HO = new List <List <double> >();
RandomList = new List <int>();
eta = pEta;
Bias = pBias;
epochs = pEpochs;
if (readfile)
{
ReadWeightFromFile();
}
else
{
Generate_Random_Weights();
}
alpha = 1;
confusionMatrix = new int[4, 4] {
{ 0, 0, 0, 0 }, { 0, 0, 0, 0 }, { 0, 0, 0, 0 }, { 0, 0, 0, 0 }
};
}