public static NeuralDataSet generateTraining() {
NeuralDataSet result = new BasicNeuralDataSet();
for (int i = 0; i < DIGITS.length; i++) {
BasicNeuralData ideal = new BasicNeuralData(DIGITS.length);
// setup input
NeuralData input = image2data(DIGITS[i]);
// setup ideal
for (int j = 0; j < DIGITS.length; j++) {
if (j == i) ideal.setData(j, 1);
else ideal.setData(j, -1);
}
// add training element
result.add(input, ideal);
}
return result;
}
/** Run method for the background training thread. */
public void run() {
try {
final int inputNeuron = OCR.DOWNSAMPLE_HEIGHT * OCR.DOWNSAMPLE_WIDTH;
final int outputNeuron = this.letterListModel.size();
final NeuralDataSet trainingSet = new BasicNeuralDataSet();
for (int t = 0; t < this.letterListModel.size(); t++) {
final NeuralData item = new BasicNeuralData(inputNeuron);
int idx = 0;
final SampleData ds = (SampleData) this.letterListModel.getElementAt(t);
for (int y = 0; y < ds.getHeight(); y++) {
for (int x = 0; x < ds.getWidth(); x++) {
item.setData(idx++, ds.getData(x, y) ? .5 : -.5);
}
}
trainingSet.add(new BasicNeuralDataPair(item, null));
}
this.net = new BasicNetwork();
this.net.addLayer(new BasicLayer(new ActivationLinear(), false, inputNeuron));
this.net.addLayer(new BasicLayer(new ActivationLinear(), false, outputNeuron));
this.net.getStructure().finalizeStructure();
this.net.reset();
final CompetitiveTraining train =
new CompetitiveTraining(
this.net, 0.25, trainingSet, new NeighborhoodGaussian(new GaussianFunction(0, 1, 2)));
int tries = 1;
while (!this.halt) {
train.iteration();
update(tries++, train.getError());
}
markStopped();
this.halt = false;
} catch (final Exception e) {
e.printStackTrace();
JOptionPane.showMessageDialog(this, "Error: " + e, "Training", JOptionPane.ERROR_MESSAGE);
}
}
