weka libsvm one-class svm

1.pom.xml引入依赖

<dependency>
      <groupId>nz.ac.waikato.cms.weka</groupId>
      <artifactId>LibSVM</artifactId>
      <version>1.0.10</version>
    </dependency>

2.样例代码

import weka.classifiers.functions.LibSVM;
import weka.core.*;
import weka.filters.Filter;
import weka.filters.unsupervised.attribute.NumericToNominal;

import java.util.ArrayList;
import java.util.Arrays;

public class OneClassSVMExample {
    public static void main(String[] args) throws Exception {
        // Create the attributes
        Attribute attribute1 = new Attribute("att1");
        Attribute attribute2 = new Attribute("att2");
        Attribute classAttribute = new Attribute("class");

        // Create the instances
        Instances trainData = new Instances("trainData", new ArrayList<>(Arrays.asList(attribute1, attribute2, classAttribute)), 0);
        Instances testData = new Instances("testData", new ArrayList<>(Arrays.asList(attribute1, attribute2, classAttribute)), 0);

        // Add the training instances
        Instance trainInstance1 = new DenseInstance(1.0, new double[]{1.0, 2.0, 1.0});
        Instance trainInstance2 = new DenseInstance(1.0, new double[]{2.0, 3.0, 1.0});
        Instance trainInstance3 = new DenseInstance(1.0, new double[]{3.0, 4.0, 1.0});
        trainData.add(trainInstance1);
        trainData.add(trainInstance2);
        trainData.add(trainInstance3);

        // Add the testing instances
        Instance testInstance1 = new DenseInstance(1.0, new double[]{1.5, 2.5, 0.0});
        Instance testInstance2 = new DenseInstance(1.0, new double[]{121.5, 131.5, 233.0});
        Instance testInstance3 = new DenseInstance(1.0, new double[]{2.5, 2.5, 1.0});
        testData.add(testInstance1);
        testData.add(testInstance2);
        testData.add(testInstance3);

        // Convert the numeric class attribute to a nominal attribute
        NumericToNominal convert = new NumericToNominal();
        convert.setAttributeIndices("last");
        convert.setInputFormat(trainData);
        trainData = Filter.useFilter(trainData, convert);

        // Set the class index
        trainData.setClassIndex(trainData.numAttributes() - 1);
        testData.setClassIndex(testData.numAttributes() - 1);

        // Create and build the OneClassSVM model
        LibSVM svm = new LibSVM();
        String[] options = new String[6];
        options[0] = "-S";
        options[1] = "2";
        options[2] = "-K";
        options[3] = "2";
        options[4] = "-G";
        options[5] = "0.05";
        svm.setOptions(options);

        svm.buildClassifier(trainData);

        // Classify the test instances
        for (Instance instance : testData) {
            double score = svm.classifyInstance(instance);
            System.out.println(score);
        }
    }
}