Encog Neural network validation/Testing

Question

I have implemented a neural network using encog library as below,

MLDataSet trainingSet = new BasicMLDataSet(XOR_INPUT, XOR_IDEAL);

    final Propagation  train =  new Backpropagation(network, trainingSet);
    int epoch = 1;
    do {
        train.iteration();
        System.out.println("Epoch #" + epoch + 
                " Error:" + train.getError());
                epoch++;

    } while (train.getError() < 0.009);

    double e = network.calculateError(trainingSet);
    System.out.println("Network trained to error :" + e);
    System.out.println("Saving Network");


    EncogDirectoryPersistence.saveObject(new File(FILENAME), network);
}


public void loadAndEvaluate(){
    System.out.println("Loading Network");
    BasicNetwork network = (BasicNetwork) EncogDirectoryPersistence.loadObject(new File(FILENAME));

    BasicMLDataSet trainingSet = new BasicMLDataSet(XOR_INPUT,XOR_IDEAL);

    double e = network.calculateError(trainingSet);

    System.out.println("Loaded network's error is (should be the same as above ):" + e);

}

This outputs the error. But i want to test this with custom data and check if the output given for a set of data is a

Answer 1

I see that you are following one of the persistence example. To obtain outputs for some input, use the "compute" function. As an example:

    double[] output = new double[1];
    network.compute(new double[]{1.0, 1.0}, output);
    System.out.println("Network output: " + output[0] + " (should be close to 0.0)");

Here's the java user guide. It's quite helpful.