minnd.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

    // The first K nearest neighbours' predictions */
  double[] predict = new double[m_NumClasses];
  for(int h=0; h < predict.length; h++)
    predict[h] = 0;
  ex = cleanse(ex);

  if(ex.relationalValue(1).numInstances() == 0){
    if (getDebug())
      System.out.println("???Whole exemplar falls into ambiguous area!");
    return 1.0;                          // Bias towards positive class
  }

  double[] mean = new double[m_Dimension];	
  for (int i=0; i<m_Dimension; i++)
    mean [i]=ex.relationalValue(1).meanOrMode(i);

  // Avoid zero sigma
  for(int h=0; h < var.length; h++){
    if(Utils.eq(var[h],0.0))
      var[h] = m_ZERO;
  }	

  for(int i=0; i < m_Class.length; i++){
    if(m_ValidM[i] != null)
      kullback[i] = kullback(mean, m_ValidM[i], var, m_Variance[i], i);
    else
      kullback[i] = Double.POSITIVE_INFINITY;
  }

  for(int j=0; j < m_Neighbour; j++){
    int pos = Utils.minIndex(kullback);
    predict[(int)m_Class[pos]] += m_Weights[pos];	   
    kullback[pos] = Double.POSITIVE_INFINITY;
  }	

  if (getDebug())
    System.out.println("???There are still some unambiguous instances in this exemplar! Predicted as: "+Utils.maxIndex(predict));
  return (double)Utils.maxIndex(predict);	
  } 

  /**
   * Cleanse the given exemplar according to the valid and noise data
   * statistics
   *
   * @param before the given exemplar
   * @return the processed exemplar
   * @throws Exception if the returned exemplar is wrong 
   */
  public Instance cleanse(Instance before) throws Exception{

    Instances insts = before.relationalValue(1).stringFreeStructure();
    Instance after = new Instance (before.numAttributes());
    after.setDataset(m_Attributes);

    for(int g=0; g < before.relationalValue(1).numInstances(); g++){
      Instance datum = before.relationalValue(1).instance(g);
      double[] minNoiDists = new double[m_Choose];
      double[] minValDists = new double[m_Choose];
      int noiseCount = 0, validCount = 0;
      double[] nDist = new double[m_Mean.length]; 
      double[] vDist = new double[m_Mean.length]; 

      for(int h=0; h < m_Mean.length; h++){
        if(m_ValidM[h] == null)
          vDist[h] = Double.POSITIVE_INFINITY;
        else
          vDist[h] = distance(datum, m_ValidM[h], m_ValidV[h], h);

        if(m_NoiseM[h] == null)
          nDist[h] = Double.POSITIVE_INFINITY;
        else
          nDist[h] = distance(datum, m_NoiseM[h], m_NoiseV[h], h);
      }

      for(int k=0; k < m_Choose; k++){
        int pos = Utils.minIndex(vDist);
        minValDists[k] = vDist[pos];
        vDist[pos] = Double.POSITIVE_INFINITY;
        pos = Utils.minIndex(nDist);
        minNoiDists[k] = nDist[pos];
        nDist[pos] = Double.POSITIVE_INFINITY;
      }

      int x = 0,y = 0;
      while((x+y) < m_Choose){
        if(minValDists[x] <= minNoiDists[y]){
          validCount++;
          x++;
        }
        else{
          noiseCount++;
          y++;
        }
      }
      if(x >= y)
        insts.add (datum);

    }

    after.setValue(0, before.value( 0));
    after.setValue(1, after.attribute(1).addRelation(insts));
    after.setValue(2, before.value( 2));

    return after;
  }    

  /**
   * This function calculates the Kullback Leibler distance between
   * two normal distributions.  This distance is always positive. 
   * Kullback Leibler distance = integral{f(X)ln(f(X)/g(X))}
   * Note that X is a vector.  Since we assume dimensions are independent
   * f(X)(g(X) the same) is actually the product of normal density
   * functions of each dimensions.  Also note that it should be log2
   * instead of (ln) in the formula, but we use (ln) simply for computational
   * convenience.
   *
   * The result is as follows, suppose there are P dimensions, and f(X)
   * is the first distribution and g(X) is the second:
   * Kullback = sum[1..P](ln(SIGMA2/SIGMA1)) +
   *            sum[1..P](SIGMA1^2 / (2*(SIGMA2^2))) +
   *            sum[1..P]((MU1-MU2)^2 / (2*(SIGMA2^2))) -
   *            P/2
   *
   * @param mu1 mu of the first normal distribution
   * @param mu2 mu of the second normal distribution 
   * @param var1 variance(SIGMA^2) of the first normal distribution
   * @param var2 variance(SIGMA^2) of the second normal distribution
   * @return the Kullback distance of two distributions
   */
  public double kullback(double[] mu1, double[] mu2,
      double[] var1, double[] var2, int pos){
    int p = mu1.length;
    double result = 0;

    for(int y=0; y < p; y++){
      if((Utils.gr(var1[y], 0)) && (Utils.gr(var2[y], 0))){
        result +=  
          ((Math.log(Math.sqrt(var2[y]/var1[y]))) +
           (var1[y] / (2.0*var2[y])) + 
           (m_Change[pos][y] * (mu1[y]-mu2[y])*(mu1[y]-mu2[y]) / (2.0*var2[y])) -
           0.5);
      }
    }

    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(
          "\tSet number of nearest neighbour for prediction\n"
          + "\t(default 1)",
          "K", 1, "-K <number of neighbours>"));
    
    result.addElement(new Option(
          "\tSet number of nearest neighbour for cleansing the training data\n"
          + "\t(default 1)",
          "S", 1, "-S <number of neighbours>"));
    
    result.addElement(new Option(
          "\tSet number of nearest neighbour for cleansing the testing data\n"
          + "\t(default 1)",
          "E", 1, "-E <number of neighbours>"));

    return result.elements();
  }

  /**
   * Parses a given list of options. <p/>
   * 
   <!-- options-start -->
   * Valid options are: <p/>
   * 
   * <pre> -K &lt;number of neighbours&gt;
   *  Set number of nearest neighbour for prediction
   *  (default 1)</pre>
   * 
   * <pre> -S &lt;number of neighbours&gt;
   *  Set number of nearest neighbour for cleansing the training data
   *  (default 1)</pre>
   * 
   * <pre> -E &lt;number of neighbours&gt;
   *  Set number of nearest neighbour for cleansing the testing data
   *  (default 1)</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 numNeighbourString = Utils.getOption('K', options);
    if (numNeighbourString.length() != 0) 
      setNumNeighbours(Integer.parseInt(numNeighbourString));
    else 
      setNumNeighbours(1);

    numNeighbourString = Utils.getOption('S', options);
    if (numNeighbourString.length() != 0) 
      setNumTrainingNoises(Integer.parseInt(numNeighbourString));
    else 
      setNumTrainingNoises(1);

    numNeighbourString = Utils.getOption('E', options);
    if (numNeighbourString.length() != 0) 
      setNumTestingNoises(Integer.parseInt(numNeighbourString));
    else 
      setNumTestingNoises(1);
  }

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

    if (getDebug())
      result.add("-D");
    
    result.add("-K");
    result.add("" + getNumNeighbours());
    
    result.add("-S");
    result.add("" + getNumTrainingNoises());
    
    result.add("-E");
    result.add("" + getNumTestingNoises());

    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 numNeighboursTipText() {
    return "The number of nearest neighbours to the estimate the class prediction of test bags.";
  }

  /**
   * Sets the number of nearest neighbours to estimate
   * the class prediction of tests bags
   * @param numNeighbour the number of citers
   */
  public void setNumNeighbours(int numNeighbour){
    m_Neighbour = numNeighbour;
  }

  /**
   * Returns the number of nearest neighbours to estimate
   * the class prediction of tests bags
   * @return the number of neighbours
   */
  public int getNumNeighbours(){
    return m_Neighbour;
  }

  /**
   * Returns the tip text for this property
   *
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String numTrainingNoisesTipText() {
    return "The number of nearest neighbour instances in the selection of noises in the training data.";
  }

  /**
   * Sets the number of nearest neighbour instances in the 
   * selection of noises in the training data
   * 
   * @param numTraining the number of noises in training data 
   */
  public void setNumTrainingNoises (int numTraining){
    m_Select = numTraining;
  }

  /**
   * Returns the number of nearest neighbour instances in the 
   * selection of noises in the training data
   * 
   * @return the number of noises in training data
   */
  public int getNumTrainingNoises(){
    return m_Select;
  }

  /**
   * Returns the tip text for this property
   *
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String numTestingNoisesTipText() {
    return "The number of nearest neighbour instances in the selection of noises in the test data.";
  }

  /**
   * Returns The number of nearest neighbour instances in the 
   * selection of noises in the test data 
   * @return the number of noises in test data
   */
  public int getNumTestingNoises(){
    return m_Choose;
  }

  /**
   * Sets The number of nearest neighbour exemplars in the 
   * selection of noises in the test data 
   * @param numTesting the number of noises in test data
   */
  public void setNumTestingNoises (int numTesting){
    m_Choose = numTesting;
  }

  /**
   * Main method for testing.
   *
   * @param args the options for the classifier
   */
  public static void main(String[] args) {	
    runClassifier(new MINND(), args);
  }
}