miboost.java

代码是一个分类器的实现,其中使用了部分weka的源代码。可以将项目导入eclipse运行

/*
 *    This program is free software; you can redistribute it and/or modify
 *    it under the terms of the GNU General Public License as published by
 *    the Free Software Foundation; either version 2 of the License, or
 *    (at your option) any later version.
 *
 *    This program is distributed in the hope that it will be useful,
 *    but WITHOUT ANY WARRANTY; without even the implied warranty of
 *    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *    GNU General Public License for more details.
 *
 *    You should have received a copy of the GNU General Public License
 *    along with this program; if not, write to the Free Software
 *    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */

/*
 * MIBoost.java
 * Copyright (C) 2005 University of Waikato, Hamilton, New Zealand
 *
 */

package weka.classifiers.mi;

import weka.classifiers.Classifier;
import weka.classifiers.SingleClassifierEnhancer;
import weka.core.Capabilities;
import weka.core.Instance;
import weka.core.Instances;
import weka.core.MultiInstanceCapabilitiesHandler;
import weka.core.Optimization;
import weka.core.Option;
import weka.core.OptionHandler;
import weka.core.TechnicalInformation;
import weka.core.TechnicalInformationHandler;
import weka.core.Utils;
import weka.core.WeightedInstancesHandler;
import weka.core.Capabilities.Capability;
import weka.core.TechnicalInformation.Field;
import weka.core.TechnicalInformation.Type;
import weka.filters.Filter;
import weka.filters.unsupervised.attribute.Discretize;
import weka.filters.unsupervised.attribute.MultiInstanceToPropositional;

import java.util.Enumeration;
import java.util.Vector;

/**
 <!-- globalinfo-start -->
 * MI AdaBoost method, considers the geometric mean of posterior of instances inside a bag (arithmatic mean of log-posterior) and the expectation for a bag is taken inside the loss function.<br/>
 * <br/>
 * For more information about Adaboost, see:<br/>
 * <br/>
 * Yoav Freund, Robert E. Schapire: Experiments with a new boosting algorithm. In: Thirteenth International Conference on Machine Learning, San Francisco, 148-156, 1996.
 * <p/>
 <!-- globalinfo-end -->
 *
 <!-- technical-bibtex-start -->
 * BibTeX:
 * <pre>
 * &#64;inproceedings{Freund1996,
 *    address = {San Francisco},
 *    author = {Yoav Freund and Robert E. Schapire},
 *    booktitle = {Thirteenth International Conference on Machine Learning},
 *    pages = {148-156},
 *    publisher = {Morgan Kaufmann},
 *    title = {Experiments with a new boosting algorithm},
 *    year = {1996}
 * }
 * </pre>
 * <p/>
 <!-- technical-bibtex-end -->
 *
 <!-- options-start -->
 * Valid options are: <p/>
 * 
 * <pre> -D
 *  Turn on debugging output.</pre>
 * 
 * <pre> -B &lt;num&gt;
 *  The number of bins in discretization
 *  (default 0, no discretization)</pre>
 * 
 * <pre> -R &lt;num&gt;
 *  Maximum number of boost iterations.
 *  (default 10)</pre>
 * 
 * <pre> -W &lt;class name&gt;
 *  Full name of classifier to boost.
 *  eg: weka.classifiers.bayes.NaiveBayes</pre>
 * 
 * <pre> -D
 *  If set, classifier is run in debug mode and
 *  may output additional info to the console</pre>
 * 
 <!-- options-end -->
 *
 * @author Eibe Frank ([e-mail])
 * @author Xin Xu ([e-mail])
 * @version $Revision: 1.5 $ 
 */
public class MIBoost 
  extends SingleClassifierEnhancer
  implements OptionHandler, MultiInstanceCapabilitiesHandler,
             TechnicalInformationHandler {

  /** for serialization */
  static final long serialVersionUID = -3808427225599279539L;
  
  /** the models for the iterations */
  protected Classifier[] m_Models;

  /** The number of the class labels */
  protected int m_NumClasses;

  /** Class labels for each bag */
  protected int[] m_Classes;

  /** attributes name for the new dataset used to build the model  */
  protected Instances m_Attributes;

  /** Number of iterations */   
  private int m_NumIterations = 100;

  /** Voting weights of models */ 
  protected double[] m_Beta;

  /** the maximum number of boost iterations */
  protected int m_MaxIterations = 10;

  /** the number of discretization bins */
  protected int m_DiscretizeBin = 0;

  /** filter used for discretization */
  protected Discretize m_Filter = null;

  /** filter used to convert the MI dataset into single-instance dataset */
  protected MultiInstanceToPropositional m_ConvertToSI = new MultiInstanceToPropositional();

  /**
   * Returns a string describing this filter
   *
   * @return a description of the filter suitable for
   * displaying in the explorer/experimenter gui
   */
  public String globalInfo() {
    return 
        "MI AdaBoost method, considers the geometric mean of posterior "
      + "of instances inside a bag (arithmatic mean of log-posterior) and "
      + "the expectation for a bag is taken inside the loss function.\n\n"
      + "For more information about Adaboost, see:\n\n"
      + getTechnicalInformation().toString();
  }

  /**
   * Returns an instance of a TechnicalInformation object, containing 
   * detailed information about the technical background of this class,
   * e.g., paper reference or book this class is based on.
   * 
   * @return the technical information about this class
   */
  public TechnicalInformation getTechnicalInformation() {
    TechnicalInformation 	result;
    
    result = new TechnicalInformation(Type.INPROCEEDINGS);
    result.setValue(Field.AUTHOR, "Yoav Freund and Robert E. Schapire");
    result.setValue(Field.TITLE, "Experiments with a new boosting algorithm");
    result.setValue(Field.BOOKTITLE, "Thirteenth International Conference on Machine Learning");
    result.setValue(Field.YEAR, "1996");
    result.setValue(Field.PAGES, "148-156");
    result.setValue(Field.PUBLISHER, "Morgan Kaufmann");
    result.setValue(Field.ADDRESS, "San Francisco");
    
    return result;
  }

  /**
   * Returns an enumeration describing the available options
   *
   * @return an enumeration of all the available options
   */
  public Enumeration listOptions() {
    Vector result = new Vector();

    result.addElement(new Option(
          "\tTurn on debugging output.",
          "D", 0, "-D"));

    result.addElement(new Option(
          "\tThe number of bins in discretization\n"
          + "\t(default 0, no discretization)",
          "B", 1, "-B <num>"));	

    result.addElement(new Option(
          "\tMaximum number of boost iterations.\n"
          + "\t(default 10)",
          "R", 1, "-R <num>"));	

    result.addElement(new Option(
          "\tFull name of classifier to boost.\n"
          + "\teg: weka.classifiers.bayes.NaiveBayes",
          "W", 1, "-W <class name>"));

    Enumeration enu = ((OptionHandler)m_Classifier).listOptions();
    while (enu.hasMoreElements()) {
      result.addElement(enu.nextElement());
    }

    return result.elements();
  }

  /**
   * Parses a given list of options. <p/>
   * 
   <!-- options-start -->
   * Valid options are: <p/>
   * 
   * <pre> -D
   *  Turn on debugging output.</pre>
   * 
   * <pre> -B &lt;num&gt;
   *  The number of bins in discretization
   *  (default 0, no discretization)</pre>
   * 
   * <pre> -R &lt;num&gt;
   *  Maximum number of boost iterations.
   *  (default 10)</pre>
   * 
   * <pre> -W &lt;class name&gt;
   *  Full name of classifier to boost.
   *  eg: weka.classifiers.bayes.NaiveBayes</pre>
   * 
   * <pre> -D
   *  If set, classifier is run in debug mode and
   *  may output additional info to the console</pre>
   * 
   <!-- options-end -->
   *
   * @param options the list of options as an array of strings
   * @throws Exception if an option is not supported
   */
  public void setOptions(String[] options) throws Exception {
    setDebug(Utils.getFlag('D', options));

    String bin = Utils.getOption('B', options);
    if (bin.length() != 0) {
      setDiscretizeBin(Integer.parseInt(bin));
    } else {
      setDiscretizeBin(0);
    }

    String boostIterations = Utils.getOption('R', options);
    if (boostIterations.length() != 0) {
      setMaxIterations(Integer.parseInt(boostIterations));
    } else {
      setMaxIterations(10);
    }

    super.setOptions(options);
  }

  /**
   * Gets the current settings of the classifier.
   *
   * @return an array of strings suitable for passing to setOptions
   */
  public String[] getOptions() {
    Vector        result;
    String[]      options;
    int           i;
    
    result  = new Vector();

    result.add("-R");
    result.add("" + getMaxIterations());

    result.add("-B");
    result.add("" + getDiscretizeBin());

    options = super.getOptions();
    for (i = 0; i < options.length; i++)
      result.add(options[i]);

    return (String[]) result.toArray(new String[result.size()]);
  }

  /**
   * Returns the tip text for this property
   *
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String maxIterationsTipText() {
    return "The maximum number of boost iterations.";
  }

  /**
   * Set the maximum number of boost iterations
   *
   * @param maxIterations the maximum number of boost iterations
   */
  public void setMaxIterations(int maxIterations) {	
    m_MaxIterations = maxIterations;
  }

  /**
   * Get the maximum number of boost iterations
   *
   * @return the maximum number of boost iterations
   */
  public int getMaxIterations() {

    return m_MaxIterations;
  }

  /**
   * Returns the tip text for this property
   *
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String discretizeBinTipText() {
    return "The number of bins in discretization.";
  }

  /**
   * Set the number of bins in discretization
   *
   * @param bin the number of bins in discretization
   */
  public void setDiscretizeBin(int bin) {	
    m_DiscretizeBin = bin;
  }

  /**
   * Get the number of bins in discretization
   *
   * @return the number of bins in discretization
   */
  public int getDiscretizeBin() {	
    return m_DiscretizeBin;
  }

  private class OptEng