Example of using JLibSVM

App.java

package org.joeltong.testJLibSvm;

import edu.berkeley.compbio.jlibsvm.ImmutableSvmParameter;
import edu.berkeley.compbio.jlibsvm.ImmutableSvmParameterGrid;
import edu.berkeley.compbio.jlibsvm.binary.BinaryModel;
import edu.berkeley.compbio.jlibsvm.binary.C_SVC;
import edu.berkeley.compbio.jlibsvm.binary.MutableBinaryClassificationProblemImpl;
import edu.berkeley.compbio.jlibsvm.kernel.LinearKernel;
import edu.berkeley.compbio.jlibsvm.util.SparseVector;

import java.util.HashSet;

/**
 * This is an example of how to use the JLibSVM library.
 * <p/>
 * It is translated from the Clojure code at
 * http://lilyx.net/2011/07/02/using-svm-support-vector-machine-from-clojure/
 * <p/>
 * The SVM is trained with 2 labels, each with 1 data entry as a proof-of-concept
 * and for simplicity.
 * <p/>
 * Putting in features characteristic of the label yields the correct result.
 * <p/>
 * Created by joel on 7/7/15.
 */
public class App {

  public App() {
    // create a new SVM implementation in the C SVC style.
    C_SVC svm = new C_SVC();
    // build parameters
    ImmutableSvmParameterGrid.Builder builder = ImmutableSvmParameterGrid.builder();

    // create training parameters ------------
    HashSet<Float> cSet;
    HashSet<LinearKernel> kernelSet;

    cSet = new HashSet<Float>();
    cSet.add(1.0f);

    kernelSet = new HashSet<LinearKernel>();
    kernelSet.add(new LinearKernel());

    // configure finetuning parameters
    builder.eps = 0.001f; // epsilon
    builder.Cset = cSet; // C values used
    builder.kernelSet = kernelSet; //Kernel used

    ImmutableSvmParameter params = builder.build();
    // / create training parameters ------------

    // create problem -------------------
    SparseVector x1 = generateFeatures(new float[]{0.1f, 0.2f});
    SparseVector x2 = generateFeatures(new float[]{0.8f, 0.9f});
    MutableBinaryClassificationProblemImpl problem
            = new MutableBinaryClassificationProblemImpl(String.class, 2);
    problem.addExample(x1, 1); // label with 1
    problem.addExample(x2, -1); // label with -1
    // / create problem -------------------

    // train ------------------------
    BinaryModel model = svm.train(problem, params);
    // / train ------------------------

    // predict -------------------------
    SparseVector xTest = generateFeatures(new float[]{0.2f, 0.1f});
    int predictedLabel = (Integer) model.predictLabel(xTest);
    System.out.println("predicted:" + predictedLabel);
    // / predict -------------------------


  }

  /**
   * Helper function to generate a single featureset.
   * @param floats
   * @return
   */
  private SparseVector generateFeatures(float[] floats) {
    SparseVector sparseVector = new SparseVector(floats.length);
    int[] indices = new int[2];
    for (int i = 0; i < floats.length; i++) {
      indices[i] = new Integer(i);
    }
    sparseVector.indexes = indices;
    sparseVector.values = floats;
    return sparseVector;
  }
}

I've searched for such an example! Thanks a lot!