实战教程：利用Java Encog类的典型应用示例

实例1: setup

import org.encog.Encog; //导入依赖的package包/类
/**
 * Setup to train the SVM.
 */
private void setup() {

	this.currentConst = this.constBegin;
	this.currentGamma = this.gammaBegin;
	this.bestError = Double.POSITIVE_INFINITY;
	this.isSetup = true;
	
	if( this.currentGamma<=0 || this.currentGamma<Encog.DEFAULT_DOUBLE_EQUAL ) {
		throw new EncogError("SVM search training cannot use a gamma value less than zero.");
	}
	
	if( this.currentConst<=0 || this.currentConst<Encog.DEFAULT_DOUBLE_EQUAL ) {
		throw new EncogError("SVM search training cannot use a const value less than zero.");
	}
	
	if( this.gammaStep<0 ) {
		throw new EncogError("SVM search gamma step cannot use a const value less than zero.");
	}
	
	if( this.constStep<0 ) {
		throw new EncogError("SVM search const step cannot use a const value less than zero.");
	}
}
 

实例2: main

import org.encog.Encog; //导入依赖的package包/类
public static void main(final String args[]) {

        MLDataSet trainingSet = new BasicMLDataSet(XOR_INPUT, XOR_IDEAL);
        NEATPopulation pop = new NEATPopulation(2,1,1000);
        pop.setInitialConnectionDensity(1.0);// not required, but speeds training
        pop.reset();

        CalculateScore score = new TrainingSetScore(trainingSet);
        // train the neural network

        final EvolutionaryAlgorithm train = NEATUtil.constructNEATTrainer(pop,score);

        do {
            train.iteration();
            System.out.println("Epoch #" + train.getIteration() + " Error:" + train.getError()+ ", Species:" + pop.getSpecies().size());
        } while(train.getError() > 0.01);

        NEATNetwork network = (NEATNetwork)train.getCODEC().decode(train.getBestGenome());

        // test the neural network
        System.out.println("Neural Network Results:");
        EncogUtility.evaluate(network, trainingSet);

        Encog.getInstance().shutdown();
    }
 

实例3: activationFunction

import org.encog.Encog; //导入依赖的package包/类
/**
 * {@inheritDoc}
 */
@Override
public final void activationFunction(final double[] x, final int start,
		final int size) {
	double sum = 0;
	for (int i = start; i < start + size; i++) {
		x[i] = BoundMath.exp(x[i]);
		sum += x[i];
	}
	if(Double.isNaN(sum) || sum <Encog.DEFAULT_DOUBLE_EQUAL ) {
		for (int i = start; i < start + size; i++) {
			x[i] = 1.0/size;
		}
	} else {
		for (int i = start; i < start + size; i++) {
			x[i] = x[i] / sum;
		}
	}
}
 

实例4: main

import org.encog.Encog; //导入依赖的package包/类
/**
 * The main method.
 * @param args No arguments are used.
 */
public static void main(final String args[]) {

    // create a neural network, without using a factory
    BasicNetwork network = new BasicNetwork();
    network.addLayer(new BasicLayer(null,true,2));
    network.addLayer(new BasicLayer(new ActivationSigmoid(),true,3));
    network.addLayer(new BasicLayer(new ActivationSigmoid(),false,1));
    network.getStructure().finalizeStructure();
    network.reset();

    // create training data
    MLDataSet trainingSet = new BasicMLDataSet(XOR_INPUT, XOR_IDEAL);

    // train the neural network
    final ResilientPropagation train = new ResilientPropagation(network, trainingSet);

    int epoch = 1;

    do {
        train.iteration();
        System.out.println("Epoch #" + epoch + " Error:" + train.getError());
        epoch++;
    } while(train.getError() > 0.01);
    train.finishTraining();

    // test the neural network
    System.out.println("Neural Network Results:");
    for(MLDataPair pair: trainingSet ) {
        final MLData output = network.compute(pair.getInput());
        System.out.println(pair.getInput().getData(0) + "," + pair.getInput().getData(1)
                + ", actual=" + output.getData(0) + ",ideal=" + pair.getIdeal().getData(0));
    }

    Encog.getInstance().shutdown();
}