All Packages  Class Hierarchy  This Package  Previous  Next  Index  WEKA's home

Class weka.classifiers.bayes.DiscreteEstimatorBayes 

java.lang.Object
    |
    +----weka.classifiers.bayes.DiscreteEstimatorBayes
public class DiscreteEstimatorBayes
extends java.lang.Object
implements Estimator, Scoreable
Symbolic probability estimator based on symbol counts and a prior.

Version:
$Revision: 1.4 $
Author:
Remco Bouckaert (rrb@xm.co.nz)

Constructor Index

o DiscreteEstimatorBayes(int, double)
Constructor

Method Index

o addValue(double, double)
Add a new data value to the current estimator.
o getCount(double)
Get a counts for a value
o getNumSymbols()
Gets the number of symbols this estimator operates with
o getProbability(double)
Get a probability estimate for a value
o logScore(int)
Gets the log score contribution of this distribution
o main(String[])
Main method for testing this class.
o toString()
Display a representation of this estimator

Constructor Detail

o DiscreteEstimatorBayes
public DiscreteEstimatorBayes(int nSymbols,
                              double fPrior)
          Constructor
Parameters:
nSymbols - the number of possible symbols (remember to include 0)
laplace - if true, counts will be initialised to 1

Method Detail

o addValue
public void addValue(double data,
                     double weight)
          Add a new data value to the current estimator.
Parameters:
data - the new data value
weight - the weight assigned to the data value
o getProbability
public double getProbability(double data)
          Get a probability estimate for a value
Parameters:
data - the value to estimate the probability of
Returns:
the estimated probability of the supplied value
o getCount
public double getCount(double data)
          Get a counts for a value
Parameters:
data - the value to get the counts for
Returns:
the count of the supplied value
o getNumSymbols
public int getNumSymbols()
          Gets the number of symbols this estimator operates with
Returns:
the number of estimator symbols
o logScore
public double logScore(int nType)
          Gets the log score contribution of this distribution
Parameters:
nType - score type
Returns:
the score
o toString
public java.lang.String toString()
          Display a representation of this estimator
Overrides:
toString in class java.lang.Object
o main
public static void main(java.lang.String argv[])
          Main method for testing this class.
Parameters:
argv - should contain a sequence of integers which will be treated as symbolic. 
All Packages  Class Hierarchy  This Package  Previous  Next  Index  WEKA's home