weka.classifiers.meta.Grading Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of weka-stable Show documentation
The Waikato Environment for Knowledge Analysis (WEKA), a machine learning workbench. This is the stable version. Apart from bugfixes, this version does not receive any other updates.
There is a newer version: 3.8.6
Show newest version
/*
 *    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.
 */

/*
 *    Grading.java
 *    Copyright (C) 2000 University of Waikato
 *
 */

package weka.classifiers.meta;

import weka.classifiers.Classifier;
import weka.core.Attribute;
import weka.core.FastVector;
import weka.core.Instance;
import weka.core.Instances;
import weka.core.RevisionUtils;
import weka.core.TechnicalInformation;
import weka.core.TechnicalInformationHandler;
import weka.core.Utils;
import weka.core.TechnicalInformation.Field;
import weka.core.TechnicalInformation.Type;

import java.util.Random;

/**
 
 * Implements Grading. The base classifiers are "graded".

 * 

 * For more information, see

 * 

 * A.K. Seewald, J. Fuernkranz: An Evaluation of Grading Classifiers. In: Advances in Intelligent Data Analysis: 4th International Conference, Berlin/Heidelberg/New York/Tokyo, 115-124, 2001.
 * 
 
 *
 
 * BibTeX:
 * 
 * @inproceedings{Seewald2001,
 *    address = {Berlin/Heidelberg/New York/Tokyo},
 *    author = {A.K. Seewald and J. Fuernkranz},
 *    booktitle = {Advances in Intelligent Data Analysis: 4th International Conference},
 *    editor = {F. Hoffmann et al.},
 *    pages = {115-124},
 *    publisher = {Springer},
 *    title = {An Evaluation of Grading Classifiers},
 *    year = {2001}
 * }
 * 
 * 
 
 *
 
 * Valid options are: 

 * 
 * 
 -M <scheme specification>
 *  Full name of meta classifier, followed by options.
 *  (default: "weka.classifiers.rules.Zero")
 * 
 *  -X <number of folds>
 *  Sets the number of cross-validation folds.
 * 
 *  -S <num>
 *  Random number seed.
 *  (default 1)
 * 
 *  -B <classifier specification>
 *  Full class name of classifier to include, followed
 *  by scheme options. May be specified multiple times.
 *  (default: "weka.classifiers.rules.ZeroR")
 * 
 *  -D
 *  If set, classifier is run in debug mode and
 *  may output additional info to the console
 * 
 
 *
 * @author Alexander K. Seewald ([email protected])
 * @author Eibe Frank ([email protected])
 * @version $Revision: 1.13 $ 
 */
public class Grading 
  extends Stacking
  implements TechnicalInformationHandler {

  /** for serialization */
  static final long serialVersionUID = 5207837947890081170L;
  
  /** The meta classifiers, one for each base classifier. */
  protected Classifier [] m_MetaClassifiers = new Classifier[0];

  /** InstPerClass */
  protected double [] m_InstPerClass = null;
    
  /**
   * Returns a string describing classifier
   * @return a description suitable for
   * displaying in the explorer/experimenter gui
   */
  public String globalInfo() {

    return 
        "Implements Grading. The base classifiers are \"graded\".\n\n"
      + "For more information, 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, "A.K. Seewald and J. Fuernkranz");
    result.setValue(Field.TITLE, "An Evaluation of Grading Classifiers");
    result.setValue(Field.BOOKTITLE, "Advances in Intelligent Data Analysis: 4th International Conference");
    result.setValue(Field.EDITOR, "F. Hoffmann et al.");
    result.setValue(Field.YEAR, "2001");
    result.setValue(Field.PAGES, "115-124");
    result.setValue(Field.PUBLISHER, "Springer");
    result.setValue(Field.ADDRESS, "Berlin/Heidelberg/New York/Tokyo");
    
    return result;
  }

  /**
   * Generates the meta data
   * 
   * @param newData the data to work on
   * @param random the random number generator used in the generation
   * @throws Exception if generation fails
   */
  protected void generateMetaLevel(Instances newData, Random random) 
    throws Exception {

    m_MetaFormat = metaFormat(newData);
    Instances [] metaData = new Instances[m_Classifiers.length];
    for (int i = 0; i < m_Classifiers.length; i++) {
      metaData[i] = metaFormat(newData);
    }
    for (int j = 0; j < m_NumFolds; j++) {

      Instances train = newData.trainCV(m_NumFolds, j, random);
      Instances test = newData.testCV(m_NumFolds, j);

      // Build base classifiers
      for (int i = 0; i < m_Classifiers.length; i++) {
	getClassifier(i).buildClassifier(train);
        for (int k = 0; k < test.numInstances(); k++) {
	  metaData[i].add(metaInstance(test.instance(k),i));
        }
      }
    }
        
    // calculate InstPerClass
    m_InstPerClass = new double[newData.numClasses()];
    for (int i=0; i < newData.numClasses(); i++) m_InstPerClass[i]=0.0;
    for (int i=0; i < newData.numInstances(); i++) {
      m_InstPerClass[(int)newData.instance(i).classValue()]++;
    }
    
    m_MetaClassifiers = Classifier.makeCopies(m_MetaClassifier,
					      m_Classifiers.length);

    for (int i = 0; i < m_Classifiers.length; i++) {
      m_MetaClassifiers[i].buildClassifier(metaData[i]);
    }
  }

  /**
   * Returns class probabilities for a given instance using the stacked classifier.
   * One class will always get all the probability mass (i.e. probability one).
   *
   * @param instance the instance to be classified
   * @throws Exception if instance could not be classified
   * successfully
   * @return the class distribution for the given instance
   */
  public double[] distributionForInstance(Instance instance) throws Exception {

    double maxPreds;
    int numPreds=0;
    int numClassifiers=m_Classifiers.length;
    int idxPreds;
    double [] predConfs = new double[numClassifiers];
    double [] preds;

    for (int i=0; iMaxInstPerClass) {
          MaxInstPerClass=(int)m_InstPerClass[i];
          MaxClass=i;
        }
      }
    }

    int predictedIndex;
    if (numPreds==1)
      predictedIndex = Utils.maxIndex(preds);
    else
    {
      // System.out.print("?");
      // System.out.print(instance.toString());
      // for (int i=0; imaxVal) {
	  maxVal=dist[j];
	  maxIdx=j;
	}
      }
      predConf= (instance.classValue()==maxIdx) ? 1:0;
    }
    
    values[idx]=predConf;
    metaInstance = new Instance(1, values);
    metaInstance.setDataset(m_MetaFormat);
    return metaInstance;
  }
  
  /**
   * Returns the revision string.
   * 
   * @return		the revision
   */
  public String getRevision() {
    return RevisionUtils.extract("$Revision: 1.13 $");
  }

  /**
   * Main method for testing this class.
   *
   * @param argv should contain the following arguments:
   * -t training file [-T test file] [-c class index]
   */
  public static void main(String [] argv) {
    runClassifier(new Grading(), argv);
  }
}