// default constructor
CNeuralNet::CNeuralNet(int numInputs, int numOutputs, int numHiddenLayers, int numUnitsPerHiddenLayer) :
m_numInputs(numInputs),
m_numOutputs(numOutputs),
m_numHiddenLayers(numHiddenLayers),
m_numLayers(numHiddenLayers+2),
m_numUnitsPerHiddenLayer(numUnitsPerHiddenLayer)
{
// set unit counts data
m_numUnitsPerLayer.Add(m_numInputs);
for(int i=0;i<m_numHiddenLayers;i++)
m_numUnitsPerLayer.Add(m_numUnitsPerHiddenLayer);
m_numUnitsPerLayer.Add(m_numOutputs);
// and alloc memory
GenerateNetwork();
// init settings for new network
m_fAlpha = tfDefaultAlpha;
m_fEta = tfDefaultEta;
m_fGain = tfDefaultGain;
//
m_nFileMajorVersion = m_nMajorVersion;
m_nFileMinorVersion = m_nMinorVersion;
m_numTrainingCycles = 0;
}
void main()
{
NET Net;
int ct = 0;
InitializeRandoms();
GenerateNetwork(&Net);
RandomWeights(&Net);
InitializeApplication(&Net);
do {
TrainNet(&Net, 10);
if (ct % 10 == 0) {
fprintf(f, "Epoch = %d\n",ct);
EvaluateNet(&Net);
}
ct++;
} while (ct<MAXCT);
fprintf(f, "\n");
fprintf(f, "Finial Epoch:\n");
EvaluateNet(&Net);
FinalizeApplication(&Net);
}
void main()
{
NET Net;
BOOL Stop;
REAL MinTestError;
InitializeRandoms();
GenerateNetwork(&Net);
RandomWeights(&Net);
InitializeApplication(&Net);
Stop = FALSE;
MinTestError = MAX_REAL;
do {
TrainNet(&Net, 10);
TestNet(&Net);
if (TestError < MinTestError) {
fprintf(f, " - saving Weights ...");
MinTestError = TestError;
SaveWeights(&Net);
}
else if (TestError > 1.2 * MinTestError) {
fprintf(f, " - stopping Training and restoring Weights ...");
Stop = TRUE;
RestoreWeights(&Net);
}
} while (NOT Stop);
TestNet(&Net);
EvaluateNet(&Net);
FinalizeApplication(&Net);
}
void main()
{
NET Net;
InitializeRandoms();
GenerateNetwork(&Net);
RandomWeights(&Net);
InitializeApplication(&Net);
TrainNet(&Net);
WriteNet(&Net);
FinalizeApplication(&Net);
}
void CNeuralNet::Serialize(CArchive& ar)
{
if (ar.IsLoading())
{
// load net info
ar >> m_nFileMajorVersion;
ar >> m_nFileMinorVersion;
ar >> m_numTrainingCycles;
ar >> m_numInputs;
ar >> m_numOutputs;
ar >> m_numLayers;
ar >> m_numHiddenLayers;
ar >> m_numUnitsPerHiddenLayer;
ar >> m_fAlpha;
ar >> m_fEta;
ar >> m_fGain;
// init layers data
m_numUnitsPerLayer.Add(m_numInputs);
for(int i=0;i<m_numHiddenLayers;i++)
m_numUnitsPerLayer.Add(m_numUnitsPerHiddenLayer);
m_numUnitsPerLayer.Add(m_numOutputs);
// create the net
ASSERT(m_pLayers[0] == NULL);
GenerateNetwork();
// then load the net
// iterate over the layers
for (int nLayer=1; nLayer<m_numLayers; nLayer++)
{
// iterate over the nodes
for (int i=1; i<=m_numUnitsPerLayer[nLayer]; i++)
{
// iterate over the connections to the previous layer
for(int j=0;j<m_numUnitsPerLayer[nLayer-1]+1;j++)
{
ar >> m_pLayers[nLayer]->ppfWeight[i][j];
ar >> m_pLayers[nLayer]->ppfWeightSave[i][j];
ar >> m_pLayers[nLayer]->ppfDeltaWeight[i][j];
}
}
}
}
void main()
{
NET Net;
REAL Error;
BOOL Stop;
INT n,m;
InitializeRandoms();
GenerateNetwork(&Net);
RandomWeights(&Net);
InitializeApplication(&Net);
do {
Error = 0;
Stop = TRUE;
for (n=0; n<NUM_DATA; n++) {
SimulateNet(&Net, Input[n], Output[n], FALSE, FALSE);
Error = MAX(Error, Net.Error);
Stop = Stop AND (Net.Error < Net.Epsilon);
}
Error = MAX(Error, Net.Epsilon);
printf("Training %0.0f%% completed ...\n", (Net.Epsilon / Error) * 100);
if (NOT Stop) {
for (m=0; m<10*NUM_DATA; m++) {
n = RandomEqualINT(0, NUM_DATA-1);
SimulateNet(&Net, Input[n], Output[n], TRUE, FALSE);
}
}
} while (NOT Stop);
for (n=0; n<NUM_DATA; n++) {
SimulateNet(&Net, Input[n], Output[n], FALSE, TRUE);
}
FinalizeApplication(&Net);
}
