void NNLayer::Calculate()
{
ASSERT( m_pPrevLayer != NULL );
VectorNeurons::iterator nit;
VectorConnections::iterator cit;
double dSum;
for( nit=m_Neurons.begin(); nit<m_Neurons.end(); nit++ )
{
NNNeuron& n = *(*nit); // to ease the terminology
cit = n.m_Connections.begin();
ASSERT( (*cit).WeightIndex < m_Weights.size() );
dSum = m_Weights[ (*cit).WeightIndex ]->value; // weight of the first connection is the bias; neuron is ignored
for ( cit++ ; cit<n.m_Connections.end(); cit++ )
{
ASSERT( (*cit).WeightIndex < m_Weights.size() );
ASSERT( (*cit).NeuronIndex < m_pPrevLayer->m_Neurons.size() );
dSum += ( m_Weights[ (*cit).WeightIndex ]->value ) *
( m_pPrevLayer->m_Neurons[ (*cit).NeuronIndex ]->output );
}
n.output = SIGMOID( dSum );
}
}
// Calculates the activation of each neuron for a particular input `_sample`
void Network::getActivation(std::vector<double> &_sample) {
// Input layer activations equal to input sample
for (int i=0; i < numInputs; i++) {
activation[0][i] = _sample[i];
}
// Solve hidden and output layer activations
for(int l=1; l < 3; l++) {
for(unsigned int i=0; i < input[l].size(); i++) {
input[l][i] = -1 * weights[l-1][i][0];
for(unsigned int j=0; j < input[l-1].size(); j++) {
input[l][i] += weights[l-1][i][j+1] * activation[l-1][j];
}
activation[l][i] = SIGMOID(input[l][i]);
}
}
}
