SVM using Encog in Java for beginners

Question

I am beginner in SVM. Could someone please help me to understand the concepts of SVM using Encog from the very basics?? It will be helpful with sample Java code.

Answer 1

In Encog SVM is just a classification or regression model and can be used mostly interchangably with other model types. I modified the Hello World XOR example to use it, you can see the results below.

This is a decent intro to them: http://webdoc.nyumc.org/nyumc/files/chibi/user-content/Final.pdf This is a more basic intro to modeling in general, I wrote it for neural networks, but it applies to SVM as well: http://www.heatonresearch.com/content/non-mathematical-introduction-using-neural-networks

package org.encog.examples.neural.xor;

import org.encog.Encog;
import org.encog.ml.data.MLData;
import org.encog.ml.data.MLDataPair;
import org.encog.ml.data.MLDataSet;
import org.encog.ml.data.basic.BasicMLDataSet;
import org.encog.ml.svm.SVM;
import org.encog.ml.svm.training.SVMTrain;

public class XORHelloWorld {

    /**
     * The input necessary for XOR.
     */
    public static double XOR_INPUT[][] = { { 0.0, 0.0 }, { 1.0, 0.0 },
            { 0.0, 1.0 }, { 1.0, 1.0 } };

    /**
     * The ideal data necessary for XOR.
     */
    public static double XOR_IDEAL[][] = { { 0.0 }, { 1.0 }, { 1.0 }, { 0.0 } };

    /**
     * The main method.
     * @param args No arguments are used.
     */
    public static void main(final String args[]) {

        // create a SVM for classification, change false to true for regression     
        SVM svm = new SVM(2,false);

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

        // train the SVM
        final SVMTrain train = new SVMTrain(svm, trainingSet);
        train.iteration();
        train.finishTraining();

        // test the SVM
        System.out.println("SVM Results:");
        for(MLDataPair pair: trainingSet ) {
            final MLData output = svm.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();
    }
}