costsensitiveclassifier.java

一个数据挖掘软件ALPHAMINERR的整个过程的JAVA版源代码

  /**
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String onDemandDirectoryTipText() {

    return "Sets the directory where cost files are loaded from. This option "
      + "is used when the costMatrixSource is set to \"On Demand\".";
  }

  /**
   * Returns the directory that will be searched for cost files when
   * loading on demand.
   *
   * @return The cost file search directory.
   */
  public File getOnDemandDirectory() {

    return m_OnDemandDirectory;
  }

  /**
   * Sets the directory that will be searched for cost files when
   * loading on demand.
   *
   * @param newDir The cost file search directory.
   */
  public void setOnDemandDirectory(File newDir) {

    if (newDir.isDirectory()) {
      m_OnDemandDirectory = newDir;
    } else {
      m_OnDemandDirectory = new File(newDir.getParent());
    }
    m_MatrixSource = MATRIX_ON_DEMAND;
  }

  /**
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String minimizeExpectedCostTipText() {

    return "Sets whether the minimum expected cost criteria will be used. If "
      + "this is false, the training data will be reweighted according to the "
      + "costs assigned to each class. If true, the minimum expected cost "
      + "criteria will be used.";
  }

  /**
   * Gets the value of MinimizeExpectedCost.
   *
   * @return Value of MinimizeExpectedCost.
   */
  public boolean getMinimizeExpectedCost() {
    
    return m_MinimizeExpectedCost;
  }
  
  /**
   * Set the value of MinimizeExpectedCost.
   *
   * @param newMinimizeExpectedCost Value to assign to MinimizeExpectedCost.
   */
  public void setMinimizeExpectedCost(boolean newMinimizeExpectedCost) {
    
    m_MinimizeExpectedCost = newMinimizeExpectedCost;
  }
  
  /**
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String classifierTipText() {
    return "Sets the Classifier used as the basis for "
      + "the cost-sensitive classification.";
  }

  /**
   * Sets the distribution classifier
   *
   * @param classifier the classifier with all options set.
   */
  public void setClassifier(Classifier classifier) {

    m_Classifier = classifier;
  }

  /**
   * Gets the classifier used.
   *
   * @return the classifier
   */
  public Classifier getClassifier() {

    return m_Classifier;
  }
  

  /**
   * Gets the classifier specification string, which contains the class name of
   * the classifier and any options to the classifier
   *
   * @return the classifier string.
   */
  protected String getClassifierSpec() {
    
    Classifier c = getClassifier();
    if (c instanceof OptionHandler) {
      return c.getClass().getName() + " "
	+ Utils.joinOptions(((OptionHandler)c).getOptions());
    }
    return c.getClass().getName();
  }
  
  /**
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String costMatrixTipText() {
    return "Sets the cost matrix explicitly. This matrix is used if the "
      + "costMatrixSource property is set to \"Supplied\".";
  }

  /**
   * Gets the misclassification cost matrix.
   *
   * @return the cost matrix
   */
  public CostMatrix getCostMatrix() {
    
    return m_CostMatrix;
  }
  
  /**
   * Sets the misclassification cost matrix.
   *
   * @param the cost matrix
   */
  public void setCostMatrix(CostMatrix newCostMatrix) {
    
    m_CostMatrix = newCostMatrix;
    m_MatrixSource = MATRIX_SUPPLIED;
  }
  
  /**
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String seedTipText() {
    return "Sets the random number seed when reweighting instances. Ignored "
      + "when using minimum expected cost criteria.";
  }
  
  /**
   * Set seed for resampling.
   *
   * @param seed the seed for resampling
   */
  public void setSeed(int seed) {

    m_Seed = seed;
  }

  /**
   * Get seed for resampling.
   *
   * @return the seed for resampling
   */
  public int getSeed() {

    return m_Seed;
  }


  /**
   * Builds the model of the base learner.
   *
   * @param data the training data
   * @exception Exception if the classifier could not be built successfully
   */
  public void buildClassifier(Instances data) throws Exception {

    if (m_Classifier == null) {
      throw new Exception("No base classifier has been set!");
    }
    if (!data.classAttribute().isNominal()) {
      throw new UnsupportedClassTypeException("Class attribute must be nominal!");
    }
    if (m_MatrixSource == MATRIX_ON_DEMAND) {
      String costName = data.relationName() + CostMatrix.FILE_EXTENSION;
      File costFile = new File(getOnDemandDirectory(), costName);
      if (!costFile.exists()) {
        throw new Exception("On-demand cost file doesn't exist: " + costFile);
      }
      setCostMatrix(new CostMatrix(new BufferedReader(
                                   new FileReader(costFile))));
    } else if (m_CostMatrix == null) {
      // try loading an old format cost file
      m_CostMatrix = new CostMatrix(data.numClasses());
      m_CostMatrix.readOldFormat(new BufferedReader(
			       new FileReader(m_CostFile)));
    }

    if (!m_MinimizeExpectedCost) {
      Random random = null;
      if (!(m_Classifier instanceof WeightedInstancesHandler)) {
	random = new Random(m_Seed);
      }
      data = m_CostMatrix.applyCostMatrix(data, random);
    }
    m_Classifier.buildClassifier(data);
  }

  /**
   * Returns class probabilities. When minimum expected cost approach is chosen,
   * returns probability one for class with the minimum expected misclassification
   * cost. Otherwise it returns the probability distribution returned by
   * the base classifier.
   *
   * @param instance the instance to be classified
   * @exception Exception if instance could not be classified
   * successfully */
  public double[] distributionForInstance(Instance instance) throws Exception {

    if (!m_MinimizeExpectedCost) {
      return m_Classifier.distributionForInstance(instance);
    }
    double [] pred = m_Classifier.distributionForInstance(instance);
    double [] costs = m_CostMatrix.expectedCosts(pred);
    /*
    for (int i = 0; i < pred.length; i++) {
      System.out.print(pred[i] + " ");
    }
    System.out.println();
    for (int i = 0; i < costs.length; i++) {
      System.out.print(costs[i] + " ");
    }
    System.out.println("\n");
    */

    // This is probably not ideal
    int classIndex = Utils.minIndex(costs);
    for (int i = 0; i  < pred.length; i++) {
      if (i == classIndex) {
	pred[i] = 1.0;
      } else {
	pred[i] = 0.0;
      }
    }
    return pred; 
  }

  /**
   *  Returns the type of graph this classifier
   *  represents.
   */   
  public int graphType() {
    
    if (m_Classifier instanceof Drawable)
      return ((Drawable)m_Classifier).graphType();
    else 
      return Drawable.NOT_DRAWABLE;
  }

  /**
   * Returns graph describing the classifier (if possible).
   *
   * @return the graph of the classifier in dotty format
   * @exception Exception if the classifier cannot be graphed
   */
  public String graph() throws Exception {
    
    if (m_Classifier instanceof Drawable)
      return ((Drawable)m_Classifier).graph();
    else throw new Exception("Classifier: " + getClassifierSpec()
			     + " cannot be graphed");
  }

  /**
   * Output a representation of this classifier
   */
  public String toString() {

    if (m_Classifier == null) {
      return "CostSensitiveClassifier: No model built yet.";
    }

    String result = "CostSensitiveClassifier using ";
      if (m_MinimizeExpectedCost) {
	result += "minimized expected misclasification cost\n";
      } else {
	result += "reweighted training instances\n";
      }
      result += "\n" + getClassifierSpec()
	+ "\n\nClassifier Model\n"
	+ m_Classifier.toString()
	+ "\n\nCost Matrix\n"
	+ m_CostMatrix.toString();

    return result;
  }

  /**
   * 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) {

    try {
      System.out.println(Evaluation
			 .evaluateModel(new CostSensitiveClassifier(),
					argv));
    } catch (Exception e) {
      System.err.println(e.getMessage());
    }
  }

}