Java Code Examples for weka.classifiers.functions.SMO#setKernel()

public void setupTestTunedClassifier() {         
    //setup classifier. in this example, if we wanted to tune the kernel as well, 
    //we'd have to extend smo and override the setOptions to allow specific options
    //for kernal settings... see SMO.setOptions()
    SMO svm = new SMO();
    PolyKernel p=new PolyKernel();
    p.setExponent(2);
    svm.setKernel(p);
    this.classifier = new SMO();                
    
    //setup tuner, defaults to 10foldCV grid-search
    this.tuner = new Tuner(); 
    
    //setup para space 
    int size = 13;
    double[] cs = new double[size];
    for (int i = 0; i < cs.length; i++)
        cs[i] = Math.pow(10.0, (i-size/2));
    
    this.space = new ParameterSpace();
    this.space.addParameter("C", cs);
}
 

public static void cheatOnMNIST(){
    Instances train=DatasetLoading.loadDataNullable("\\\\cmptscsvr.cmp.uea.ac.uk\\ueatsc\\Data\\LargeProblems\\MNIST\\MNIST_TRAIN");
    Instances test=DatasetLoading.loadDataNullable("\\\\cmptscsvr.cmp.uea.ac.uk\\ueatsc\\Data\\LargeProblems\\MNIST\\MNIST_TEST");
    SMO svm=new SMO();
    RBFKernel k=new RBFKernel();
    svm.setKernel(k);
    System.out.println("Data loaded ......");
    double a =ClassifierTools.singleTrainTestSplitAccuracy(svm, train, test);
    System.out.println("Default acc = "+a);
    int min=1;//These search values are used for all kernels with C. It is also used for Gamma in RBF, but not for the Polynomial exponent search
    int max=6;
    for(double c=min;c<=max;c++)
        for(double r=min;r<=max;r++){
                 svm.setC(Math.pow(2, c));
                 k.setGamma(Math.pow(2, r));
                 svm.setKernel(k);//Just in case ...
                a =ClassifierTools.singleTrainTestSplitAccuracy(svm, train, test);
                System.out.println("logC ="+c+" logGamma = "+r+" acc = "+a);
            }
    
}
 

public void configureEnsemble(){
    ensemble.setWeightingScheme(new TrainAcc(4));
    ensemble.setVotingScheme(new MajorityConfidence());
    
    Classifier[] classifiers = new Classifier[3];
    String[] classifierNames = new String[3];
    
    SMO smo = new SMO();
    smo.turnChecksOff();
    smo.setBuildLogisticModels(true);
    PolyKernel kl = new PolyKernel();
    kl.setExponent(2);
    smo.setKernel(kl);
    if (seedClassifier)
        smo.setRandomSeed((int)seed);
    classifiers[0] = smo;
    classifierNames[0] = "SVMQ";

    RandomForest r=new RandomForest();
    r.setNumTrees(500);
    if(seedClassifier)
       r.setSeed((int)seed);            
    classifiers[1] = r;
    classifierNames[1] = "RandF";
        
        
    RotationForest rf=new RotationForest();
    rf.setNumIterations(100);
    if(seedClassifier)
       rf.setSeed((int)seed);
    classifiers[2] = rf;
    classifierNames[2] = "RotF";
    
    
   ensemble.setClassifiers(classifiers, classifierNames, null);        
    
}
 

public static void main(String[] args) throws Exception {
    int seed = 0;
    
    SMO svm = new SMO();
    PolyKernel p=new PolyKernel();
    p.setExponent(2);
    svm.setKernel(p);
    svm.setRandomSeed(seed);
    
    int size = 5;
    double[] cs = new double[size];
    for (int i = 0; i < cs.length; i++)
        cs[i] = Math.pow(10.0, (i-size/2));
    
    ParameterSpace space = new ParameterSpace();
    space.addParameter("C", cs);
    
    Tuner tuner = new Tuner();
    tuner.setPathToSaveParameters("C:/Temp/TunerTests/first/");
    tuner.setSeed(seed);
    
    String dataset = "hayes-roth";
    Instances all = DatasetLoading.loadDataNullable("Z:\\Data\\UCIDelgado\\"+dataset+"\\"+dataset+".arff");
    Instances[] data = InstanceTools.resampleInstances(all, seed, 0.5);
    
    System.out.println(tuner.tune(svm, data[0], space));
}
 

/**
 * Uses the 'basic UCI' set up:
 * Comps: SVML, MLP, NN, Logistic, C4.5
 * Weight: TrainAcc(4) (train accuracies to the power 4)
 * Vote: MajorityConfidence (summing probability distributions)
 */
public final void setupDefaultSettings_NoLogistic() {
    this.ensembleName = "CAWPE-NoLogistic";
    
    this.weightingScheme = new TrainAcc(4);
    this.votingScheme = new MajorityConfidence();
    
    CrossValidationEvaluator cv = new CrossValidationEvaluator(seed, false, false, false, false); 
    cv.setNumFolds(10);
    this.trainEstimator = cv; 
    
    Classifier[] classifiers = new Classifier[4];
    String[] classifierNames = new String[4];

    SMO smo = new SMO();
    smo.turnChecksOff();
    smo.setBuildLogisticModels(true);
    PolyKernel kl = new PolyKernel();
    kl.setExponent(1);
    smo.setKernel(kl);
    smo.setRandomSeed(seed);
    classifiers[0] = smo;
    classifierNames[0] = "SVML";

    kNN k=new kNN(100);
    k.setCrossValidate(true);
    k.normalise(false);
    k.setDistanceFunction(new EuclideanDistance());
    classifiers[1] = k;
    classifierNames[1] = "NN";

    classifiers[2] = new J48();
    classifierNames[2] = "C4.5";

    classifiers[3] = new MultilayerPerceptron();
    classifierNames[3] = "MLP";

    setClassifiers(classifiers, classifierNames, null);
}
 

public final void setupAdvancedSettings() {
    this.ensembleName = "CAWPE-A";
    
    this.weightingScheme = new TrainAcc(4);
    this.votingScheme = new MajorityConfidence();
    
    CrossValidationEvaluator cv = new CrossValidationEvaluator(seed, false, false, false, false); 
    cv.setNumFolds(10);
    this.trainEstimator = cv; 

    Classifier[] classifiers = new Classifier[3];
    String[] classifierNames = new String[3];

    SMO smo = new SMO();
    smo.turnChecksOff();
    smo.setBuildLogisticModels(true);
    PolyKernel kl = new PolyKernel();
    kl.setExponent(2);
    smo.setKernel(kl);
    smo.setRandomSeed(seed);
    classifiers[0] = smo;
    classifierNames[0] = "SVMQ";
    RandomForest rf= new RandomForest();
    rf.setNumTrees(500);
    classifiers[1] = rf;
    classifierNames[1] = "RandF";
    RotationForest rotf=new RotationForest();
    rotf.setNumIterations(200);
    classifiers[2] = rotf;
    classifierNames[2] = "RotF";

    setClassifiers(classifiers, classifierNames, null);
}
 

/**
 * Initializes the CAWPE ensemble model consisting of five classifiers (SMO,
 * KNN, J48, Logistic and MLP) using a majority voting strategy. The ensemble
 * uses Weka classifiers. It refers to "Heterogeneous ensemble of standard
 * classification algorithms" (HESCA) as described in Lines, Jason & Taylor,
 * Sarah & Bagnall, Anthony. (2018). Time Series Classification with HIVE-COTE:
 * The Hierarchical Vote Collective of Transformation-Based Ensembles. ACM
 * Transactions on Knowledge Discovery from Data. 12. 1-35. 10.1145/3182382.
 *
 * @param seed
 *            Seed used within the classifiers and the majority confidence
 *            voting scheme
 * @param numFolds
 *            Number of folds used within the determination of the classifier
 *            weights for the {@link MajorityConfidenceVote}
 * @return Returns an initialized (but untrained) ensemble model.
 * @throws Exception
 *             Thrown when the initialization has failed
 */
public static Classifier provideCAWPEEnsembleModel(final int seed, final int numFolds) throws Exception {
	Classifier[] classifiers = new Classifier[5];

	Vote voter = new MajorityConfidenceVote(numFolds, seed);

	SMO smo = new SMO();
	smo.turnChecksOff();
	smo.setBuildCalibrationModels(true);
	PolyKernel kl = new PolyKernel();
	kl.setExponent(1);
	smo.setKernel(kl);
	smo.setRandomSeed(seed);
	classifiers[0] = smo;

	IBk k = new IBk(100);
	k.setCrossValidate(true);
	EuclideanDistance ed = new EuclideanDistance();
	ed.setDontNormalize(true);
	k.getNearestNeighbourSearchAlgorithm().setDistanceFunction(ed);
	classifiers[1] = k;

	J48 c45 = new J48();
	c45.setSeed(seed);
	classifiers[2] = c45;

	classifiers[3] = new Logistic();

	classifiers[4] = new MultilayerPerceptron();

	voter.setClassifiers(classifiers);
	return voter;
}
 

/**
 * Classifiers used in the HIVE COTE paper
 */    
    public void configureDefaultEnsemble(){
//HIVE_SHAPELET_SVMQ    HIVE_SHAPELET_RandF    HIVE_SHAPELET_RotF    
//HIVE_SHAPELET_NN    HIVE_SHAPELET_NB    HIVE_SHAPELET_C45    HIVE_SHAPELET_SVML   
        ensemble=new CAWPE();
        ensemble.setWeightingScheme(new TrainAcc(4));
        ensemble.setVotingScheme(new MajorityConfidence());
        Classifier[] classifiers = new Classifier[7];
        String[] classifierNames = new String[7];
        
        SMO smo = new SMO();
        smo.turnChecksOff();
        smo.setBuildLogisticModels(true);
        PolyKernel kl = new PolyKernel();
        kl.setExponent(2);
        smo.setKernel(kl);
        if (seedClassifier)
            smo.setRandomSeed((int)seed);
        classifiers[0] = smo;
        classifierNames[0] = "SVMQ";

        RandomForest r=new RandomForest();
        r.setNumTrees(500);
        if(seedClassifier)
           r.setSeed((int)seed);            
        classifiers[1] = r;
        classifierNames[1] = "RandF";
            
            
        RotationForest rf=new RotationForest();
        rf.setNumIterations(100);
        if(seedClassifier)
           rf.setSeed((int)seed);
        classifiers[2] = rf;
        classifierNames[2] = "RotF";
        IBk nn=new IBk();
        classifiers[3] = nn;
        classifierNames[3] = "NN";
        NaiveBayes nb=new NaiveBayes();
        classifiers[4] = nb;
        classifierNames[4] = "NB";
        J48 c45=new J48();
        classifiers[5] = c45;
        classifierNames[5] = "C45";
        SMO svml = new SMO();
        svml.turnChecksOff();
        svml.setBuildLogisticModels(true);
        PolyKernel k2 = new PolyKernel();
        k2.setExponent(1);
        smo.setKernel(k2);
        classifiers[6] = svml;
        classifierNames[6] = "SVML";
        ensemble.setClassifiers(classifiers, classifierNames, null);
    }
 

@Override
    public void buildClassifier(Instances train) throws Exception {
        res =new ClassifierResults();
        long t=System.currentTimeMillis();
//        if(kernelOptimise)
//            selectKernel(train);
        if(buildFromPartial){
            if(paraSpace1==null)
                setStandardParaSearchSpace();  
            if(kernel==KernelType.RBF)
                setRBFParasFromPartiallyCompleteSearch();            
//            else if(kernel==KernelType.LINEAR || kernel==KernelType.QUADRATIC)
//                setFixedPolynomialParasFromPartiallyCompleteSearch();            
                else if(kernel==KernelType.POLYNOMIAL)
                    setPolynomialParasFromPartiallyCompleteSearch();            
        }
        else if(tuneParameters){
            if(paraSpace1==null)
                setStandardParaSearchSpace();
            if(buildFromFile){
                throw new Exception("Build from file in TunedSVM Not implemented yet");
            }else{
                if(kernel==KernelType.RBF)
                    tuneRBF(train); //Tunes two parameters
                else if(kernel==KernelType.LINEAR || kernel==KernelType.QUADRATIC)
                    tuneCForFixedPolynomial(train);//Tunes one parameter
                else if(kernel==KernelType.POLYNOMIAL)
                    tunePolynomial(train);
            }
        }
/*If there is no parameter search, then there is no train CV available.        
this gives the option of finding one using 10xCV  
*/        
        else if(findTrainAcc){
            int folds=10;
            if(folds>train.numInstances())
                folds=train.numInstances();
            SMO model = new SMO();
            model.setKernel(this.m_kernel);
            model.setC(this.getC());
            model.setBuildLogisticModels(true);
            model.setRandomSeed(seed);
            CrossValidationEvaluator cv = new CrossValidationEvaluator();
            cv.setSeed(seed); //trying to mimick old seeding behaviour below
            cv.setNumFolds(folds);
            cv.buildFolds(train);
            res = cv.crossValidateWithStats(model, train);
      }        
        
        
        
//If both kernelOptimise and tuneParameters are false, it just builds and SVM        
//With whatever the parameters are set to        
        super.buildClassifier(train);
        
        if(saveEachParaAcc)
            res.setBuildTime(combinedBuildTime);
        else
            res.setBuildTime(System.currentTimeMillis()-t);
        if(trainPath!=null && trainPath!=""){  //Save basic train results
            res.setClassifierName("TunedSVM"+kernel);
            res.setDatasetName(train.relationName());
            res.setFoldID(seed);
            res.setSplit("train");
            
            res.setParas(getParameters());
            res.writeFullResultsToFile(trainPath);
            File x=new File(trainPath);
            x.setWritable(true, false);
            
        }        
    }
 

/**
 * Uses the 'basic UCI' set up:
 * Comps: SVML, MLP, NN, Logistic, C4.5
 * Weight: TrainAcc(4) (train accuracies to the power 4)
 * Vote: MajorityConfidence (summing probability distributions)
 */
@Override //Abstract Ensemble 
public final void setupDefaultEnsembleSettings() {
    this.ensembleName = "CAWPE";
    
    this.weightingScheme = new TrainAcc(4);
    this.votingScheme = new MajorityConfidence();
    this.transform = null;
    
    CrossValidationEvaluator cv = new CrossValidationEvaluator(seed, false, false, false, false); 
    cv.setNumFolds(10);
    this.trainEstimator = cv; 

    Classifier[] classifiers = new Classifier[5];
    String[] classifierNames = new String[5];

    SMO smo = new SMO();
    smo.turnChecksOff();
    smo.setBuildLogisticModels(true);
    PolyKernel kl = new PolyKernel();
    kl.setExponent(1);
    smo.setKernel(kl);
    smo.setRandomSeed(seed);
    classifiers[0] = smo;
    classifierNames[0] = "SVML";

    kNN k=new kNN(100);
    k.setCrossValidate(true);
    k.normalise(false);
    k.setDistanceFunction(new EuclideanDistance());
    classifiers[1] = k;
    classifierNames[1] = "NN";

    classifiers[2] = new J48();
    classifierNames[2] = "C4.5";

    classifiers[3] = new Logistic();
    classifierNames[3] = "Logistic";

    classifiers[4] = new MultilayerPerceptron();
    classifierNames[4] = "MLP";
    
    setClassifiers(classifiers, classifierNames, null);
}
 

/**
     * Comps: NN, SVML, SVMQ, C4.5, NB,  RotF, RandF, BN,
     * Weight: TrainAcc
     * Vote: MajorityVote
     *
     * As used originally in ST_HESCA, COTE.
     * NOTE the original also contained Bayes Net (BN). We have removed it because the classifier crashes
     * unpredictably when discretising features (due to lack of variance in the feature, but not easily detected and
     * dealt with
     *
     */
    public final void setupOriginalHESCASettings() {
        this.ensembleName = "HESCA";
        
        this.weightingScheme = new TrainAcc();
        this.votingScheme = new MajorityVote();
        
        CrossValidationEvaluator cv = new CrossValidationEvaluator(seed, false, false, false, false); 
        cv.setNumFolds(10);
        this.trainEstimator = cv; 
        int numClassifiers=7;
        Classifier[] classifiers = new Classifier[numClassifiers];
        String[] classifierNames = new String[numClassifiers];

        kNN k=new kNN(100);
        k.setCrossValidate(true);
        k.normalise(false);
        k.setDistanceFunction(new EuclideanDistance());
        classifiers[0] = k;
        classifierNames[0] = "NN";

        classifiers[1] = new NaiveBayes();
        classifierNames[1] = "NB";

        classifiers[2] = new J48();
        classifierNames[2] = "C45";

        SMO svml = new SMO();
        svml.turnChecksOff();
        PolyKernel kl = new PolyKernel();
        kl.setExponent(1);
        svml.setKernel(kl);
        svml.setRandomSeed(seed);
        classifiers[3] = svml;
        classifierNames[3] = "SVML";

        SMO svmq =new SMO();
//Assumes no missing, all real valued and a discrete class variable
        svmq.turnChecksOff();
        PolyKernel kq = new PolyKernel();
        kq.setExponent(2);
        svmq.setKernel(kq);
        svmq.setRandomSeed(seed);
        classifiers[4] =svmq;
        classifierNames[4] = "SVMQ";

        RandomForest r=new RandomForest();
        r.setNumTrees(500);
        r.setSeed(seed);
        classifiers[5] = r;
        classifierNames[5] = "RandF";

        RotationForest rf=new RotationForest();
        rf.setNumIterations(50);
        rf.setSeed(seed);
        classifiers[6] = rf;
        classifierNames[6] = "RotF";

//        classifiers[7] = new BayesNet();
//        classifierNames[7] = "bayesNet";

        setClassifiers(classifiers, classifierNames, null);
    }
 

/**
 * Initializes the HIVE COTE ensemble consisting of 7 classifiers using a
 * majority voting strategy as described in J. Lines, S. Taylor and A. Bagnall,
 * "HIVE-COTE: The Hierarchical Vote Collective of Transformation-Based
 * Ensembles for Time Series Classification," 2016 IEEE 16th International
 * Conference on Data Mining (ICDM), Barcelona, 2016, pp. 1041-1046. doi:
 * 10.1109/ICDM.2016.0133.
 *
 * @param seed
 *            Seed used within the classifiers and the majority confidence
 *            voting scheme
 * @param numFolds
 *            Number of folds used within the determination of the classifier
 *            weights for the {@link MajorityConfidenceVote}
 * @return Returns the initialized (but untrained) HIVE COTE ensemble model.
 */
public static Classifier provideHIVECOTEEnsembleModel(final long seed) {
	Classifier[] classifier = new Classifier[7];

	Vote voter = new MajorityConfidenceVote(5, seed);

	// SMO poly2
	SMO smop = new SMO();
	smop.turnChecksOff();
	smop.setBuildCalibrationModels(true);
	PolyKernel kernel = new PolyKernel();
	kernel.setExponent(2);
	smop.setKernel(kernel);
	smop.setRandomSeed((int)seed);
	classifier[0] = smop;

	// Random Forest
	RandomForest rf = new RandomForest();
	rf.setSeed((int)seed);
	rf.setNumIterations(500);
	classifier[1] = rf;

	// Rotation forest
	RotationForest rotF = new RotationForest();
	rotF.setSeed((int)seed);
	rotF.setNumIterations(100);
	classifier[2] = rotF;

	// NN
	IBk nn = new IBk();
	classifier[3] = nn;

	// Naive Bayes
	NaiveBayes nb = new NaiveBayes();
	classifier[4] = nb;

	// C45
	J48 c45 = new J48();
	c45.setSeed((int)seed);
	classifier[5] = c45;

	// SMO linear
	SMO smol = new SMO();
	smol.turnChecksOff();
	smol.setBuildCalibrationModels(true);
	PolyKernel linearKernel = new PolyKernel();
	linearKernel.setExponent(1);
	smol.setKernel(linearKernel);
	classifier[6] = smol;

	voter.setClassifiers(classifier);
	return voter;
}