Main Page | Class Hierarchy | Class List | Class Members

kddml.Core.DataMining.Clustering.EM_DISTANCE Class Reference

Inheritance diagram for kddml.Core.DataMining.Clustering.EM_DISTANCE:

kddml.Core.DataMining.Clustering.DistributionBasedComparisonMeasure kddml.Core.DataMining.Clustering.ComparisonMeasure List of all members.

Public Member Functions

 EM_DISTANCE (double[] cluster_priors)
double[] getDistributionProbability (Cluster[] clusters, Instance instance) throws ClusteringModelException
String getFunctionName ()

Protected Member Functions

Element toXML ()

Package Attributes

double[] priors

Static Package Attributes

static final double log2 = Math.log(2)
static final double m_normConst = Math.log(Math.sqrt(2 * Math.PI))

Detailed Description

Proprietary comparison function used by EM algorithm. EM assigns a probability distribution to each instance which indicates the probability of it belonging to each of the clusters. For each cluster, a prior probability is given.

Title: KDDML

Description: Knowledge Discovery in Database Environment

Copyright: Copyright (c) 2003-2005

Company: Universita' di Pisa - Dipartimento di Informatica

Author:
Andrea Romei (romei@di.unipi.it)
Version:
2.0.16


Constructor & Destructor Documentation

kddml.Core.DataMining.Clustering.EM_DISTANCE.EM_DISTANCE double[]  cluster_priors  ) 
 

Constructor given the prior probability for each cluster.

Parameters:
cluster_priors double[]


Member Function Documentation

double [] kddml.Core.DataMining.Clustering.EM_DISTANCE.getDistributionProbability Cluster[]  clusters,
Instance  instance
throws ClusteringModelException [virtual]
 

Given an input instance, returns a distribution probability in which the i-esim array location contains the probability that the instance belongs to the cluster i.

Parameters:
clusters Cluster[]
instance Instance
Exceptions:
ClusteringModelException 
Returns:
double[] a normalized array.

Implements kddml.Core.DataMining.Clustering.DistributionBasedComparisonMeasure.

String kddml.Core.DataMining.Clustering.EM_DISTANCE.getFunctionName  )  [virtual]
 

Returns the function name as in PMML.

Returns:
String

Implements kddml.Core.DataMining.Clustering.ComparisonMeasure.

Element kddml.Core.DataMining.Clustering.EM_DISTANCE.toXML  )  [protected, virtual]
 

Returns a representation of this function as PMML element.

Returns:
Element

Implements kddml.Core.DataMining.Clustering.ComparisonMeasure.


Member Data Documentation

final double kddml.Core.DataMining.Clustering.EM_DISTANCE.log2 = Math.log(2) [static, package]
 

The natural logarithm of 2.

final double kddml.Core.DataMining.Clustering.EM_DISTANCE.m_normConst = Math.log(Math.sqrt(2 * Math.PI)) [static, package]
 

Constant for normal distribution.

double [] kddml.Core.DataMining.Clustering.EM_DISTANCE.priors [package]
 

Priors probability.


Generated on Thu Feb 23 13:04:41 2006 for kddml by  doxygen 1.4.3