Package es.uam.eps.ir.relison.links.linkprediction.supervised.classifiers

Class NaiveBayesClassifier<U>

java.lang.Object
es.uam.eps.ir.relison.links.linkprediction.supervised.classifiers.NaiveBayesClassifier<U>
All Implemented Interfaces:
Classifier<U>
public class NaiveBayesClassifier<U>
extends java.lang.Object
implements Classifier<U>
Classifier which applies the Naive Bayes method. In case the attributes are continuous, Gaussian Naive Bayes is applied to compute the scores. Note: For computing the means and variances, an incremental algorithm is used. This algorithm is documented in:

  • Field Summary

    Fields 
    Modifier and Type Field Description
    private java.util.List<java.lang.String> attributes
    List of attributes.
    private java.util.List<java.lang.Integer> classes
    List of classes.
    private java.util.List<java.lang.Double[][]> frequencies
    A list of matrices which contains the information for each attribute in the training set.
    private static int MEAN
    Constant index for the mean.
    private boolean normalize
    Indicates if attributes have to be normalized or not.
    private static int NUMCONT
    Number of parameters in .
    private int numInstances
    The number of patterns.
    private java.util.List<java.lang.Double> priori
    Times each class appears in the training set.
    private static int SIGMA
    Constant index for the variance.
    private java.util.List<es.uam.eps.ir.ranksys.core.util.Stats> stats
    The stats for each attribute.
    private boolean trained
    Indicates if the training of the classifier has been done.
    private java.util.List<FeatureType> types
    The types for each attribute.

  • Constructor Summary

    Constructors 
    Constructor Description
    NaiveBayesClassifier()
    Constructor.
    NaiveBayesClassifier​(boolean normalize)
    Constructor

  • Method Summary

    Modifier and Type Method Description
    int classify​(Instance<U> instance)
    Obtains the most likely class for a certain instance.
    double computeScore​(Instance<U> instance, int category)
    Gets the score for an individual pattern in a certain category.
    java.util.Map<java.lang.Integer,​java.lang.Double> computeScores​(Instance<U> instance)
    Computes the scores for an individual instance (once the training has been done).
    void train​(InstanceSet<U> trainSet)
    Trains the classifier.

    Methods inherited from class java.lang.Object

    clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

  • Field Details

    • trained

      private boolean trained
      Indicates if the training of the classifier has been done.

    • priori

      private java.util.List<java.lang.Double> priori
      Times each class appears in the training set.

    • frequencies

      private java.util.List<java.lang.Double[][]> frequencies
      A list of matrices which contains the information for each attribute in the training set. For each matrix, the columns represent the different classes. In the case of nominal attributes, each row represents a different value of the attributes, and each cell contains the number of different examples which share the same value of the attribute and class. In the case of continuous attributes, first row represents the mean of the sample, and second row represents the variance of the sample.

    • classes

      private java.util.List<java.lang.Integer> classes
      List of classes.

    • attributes

      private java.util.List<java.lang.String> attributes
      List of attributes.

    • MEAN

      private static final int MEAN
      Constant index for the mean.
      See Also:
      Constant Field Values

    • SIGMA

      private static final int SIGMA
      Constant index for the variance.
      See Also:
      Constant Field Values

    • NUMCONT

      private static final int NUMCONT
      Number of parameters in .
      See Also:
      Constant Field Values

    • numInstances

      private int numInstances
      The number of patterns.

    • stats

      private java.util.List<es.uam.eps.ir.ranksys.core.util.Stats> stats
      The stats for each attribute.

    • types

      private java.util.List<FeatureType> types
      The types for each attribute.

    • normalize

      private final boolean normalize
      Indicates if attributes have to be normalized or not.

  • Constructor Details

    • NaiveBayesClassifier

      public NaiveBayesClassifier()
      Constructor.

    • NaiveBayesClassifier

      public NaiveBayesClassifier​(boolean normalize)
      Constructor
      Parameters:
      normalize - indicates if attributes have to be normalized

  • Method Details

    • train

      public void train​(InstanceSet<U> trainSet)
      Description copied from interface: Classifier
      Trains the classifier.
      Specified by:
      train in interface Classifier<U>
      Parameters:
      trainSet - the training set.

    • computeScores

      public java.util.Map<java.lang.Integer,​java.lang.Double> computeScores​(Instance<U> instance)
      Description copied from interface: Classifier
      Computes the scores for an individual instance (once the training has been done).
      Specified by:
      computeScores in interface Classifier<U>
      Parameters:
      instance - the individual instance.
      Returns:
      A score for each class.

    • computeScore

      public double computeScore​(Instance<U> instance, int category)
      Description copied from interface: Classifier
      Gets the score for an individual pattern in a certain category.
      Specified by:
      computeScore in interface Classifier<U>
      Parameters:
      instance - the pattern.
      category - the class.
      Returns:
      the score.

    • classify

      public int classify​(Instance<U> instance)
      Description copied from interface: Classifier
      Obtains the most likely class for a certain instance.
      Specified by:
      classify in interface Classifier<U>
      Parameters:
      instance - the individual instance.
      Returns:
      The most likely class.