gov.sandia.cognition.learning.algorithm.clustering.cluster

Class MedoidClusterCreator<DataType>

java.lang.Object

gov.sandia.cognition.util.AbstractCloneableSerializable

gov.sandia.cognition.learning.function.distance.DefaultDivergenceFunctionContainer<DataType,DataType>

gov.sandia.cognition.learning.algorithm.clustering.cluster.MedoidClusterCreator<DataType>

Type Parameters:

DataType - The algorithm operates on a Collection of DataType, so DataType will be something like Vector or String

All Implemented Interfaces:

ClusterCreator<CentroidCluster<DataType>,DataType>, DivergenceFunctionContainer<DataType,DataType>, CloneableSerializable, java.io.Serializable, java.lang.Cloneable

@CodeReview(reviewer="Kevin R. Dixon",
            date="2008-07-22",
            changesNeeded=false,
            comments="Code generally looks fine.")
public class MedoidClusterCreator<DataType>
extends DefaultDivergenceFunctionContainer<DataType,DataType>
implements ClusterCreator<CentroidCluster<DataType>,DataType>

The MedoidClusterCreator class creates a CentroidCluster at the sample that minimizes the sum of the divergence to the objects assigned to the cluster.

Since:

2.0

Author:

Justin Basilico, Kevin R. Dixon

See Also:

Serialized Form

Field Summary

Fields inherited from class gov.sandia.cognition.learning.function.distance.DefaultDivergenceFunctionContainer

divergenceFunction

Constructor Summary

Constructors

Constructor and Description
MedoidClusterCreator() Creates a new instance of MedoidClusterCreator
MedoidClusterCreator(DivergenceFunction<? super DataType,? super DataType> divergenceFunction) Creates a new instance of MedoidClusterCreator

Method Summary

Modifier and Type	Method and Description
CentroidCluster<DataType>	createCluster(java.util.Collection<? extends DataType> members) Creates a CentroidCluster at the member that minimizes the sum of divergence between all members

Methods inherited from class gov.sandia.cognition.learning.function.distance.DefaultDivergenceFunctionContainer

clone, getDivergenceFunction, setDivergenceFunction

Methods inherited from class java.lang.Object

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

Constructor Detail

MedoidClusterCreator

public MedoidClusterCreator()

Creates a new instance of MedoidClusterCreator

MedoidClusterCreator

public MedoidClusterCreator(DivergenceFunction<? super DataType,? super DataType> divergenceFunction)

Creates a new instance of MedoidClusterCreator

Parameters:

divergenceFunction - Divergence function used to evaluate the dissimilarity between two data points

Method Detail

createCluster

public CentroidCluster<DataType> createCluster(java.util.Collection<? extends DataType> members)

Creates a CentroidCluster at the member that minimizes the sum of divergence between all members

Specified by:

createCluster in interface ClusterCreator<CentroidCluster<DataType>,DataType>

Parameters:

members - Data points that have been assigned to the cluster

Returns:

CentroidCluster that minimizes the sum of divergence between all assigned members