de.lmu.ifi.dbs.elki.algorithm.clustering.subspace

Class HiSC<V extends NumberVector>

java.lang.Object

de.lmu.ifi.dbs.elki.algorithm.AbstractAlgorithm<ClusterOrderResult<E>>

de.lmu.ifi.dbs.elki.algorithm.clustering.optics.GeneralizedOPTICS<V,HiSC.HiSCClusterOrderEntry>

de.lmu.ifi.dbs.elki.algorithm.clustering.subspace.HiSC<V>

Type Parameters:

V - the type of NumberVector handled by the algorithm

All Implemented Interfaces:

Algorithm, OPTICSTypeAlgorithm<HiSC.HiSCClusterOrderEntry>

@Title(value="Finding Hierarchies of Subspace Clusters")
 @Description(value="Algorithm for detecting hierarchies of subspace clusters.")
 @Reference(authors="E. Achtert, C. B\u00f6hm, H.-P. Kriegel, P. Kr\u00f6ger, I. M\u00fcller-Gorman, A. Zimek",
           title="Finding Hierarchies of Subspace Clusters",
           booktitle="Proc. 10th Europ. Conf. on Principles and Practice of Knowledge Discovery in Databases (PKDD\'06), Berlin, Germany, 2006",
           url="http://www.dbs.ifi.lmu.de/Publikationen/Papers/PKDD06-HiSC.pdf")
public class HiSC<V extends NumberVector>
extends GeneralizedOPTICS<V,HiSC.HiSCClusterOrderEntry>

Implementation of the HiSC algorithm, an algorithm for detecting hierarchies of subspace clusters.

Reference: E. Achtert, C. Böhm, H.-P. Kriegel, P. Kröger, I. Müller-Gorman, A. Zimek: Finding Hierarchies of Subspace Clusters.
In: Proc. 10th Europ. Conf. on Principles and Practice of Knowledge Discovery in Databases (PKDD'06), Berlin, Germany, 2006.

Author:

Elke Achtert

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	HiSC.HiSCClusterOrderEntry Cluster order entry for HiSC.
static class	HiSC.Parameterizer<V extends NumberVector> Parameterization class.

Field Summary

Fields

Modifier and Type	Field and Description
private double	alpha Holds the maximum diversion allowed.
private HiSCPreferenceVectorIndex<NumberVector>	index Instantiated index.
private IndexFactory<V,HiSCPreferenceVectorIndex<NumberVector>>	indexfactory Factory to produce
private static Logging	LOG The logger for this class.
private Relation<V>	relation Relation we are currently processing.

Fields inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.optics.GeneralizedOPTICS

MINPTS_ID

Constructor Summary

Constructors

Constructor and Description
HiSC(IndexFactory<V,HiSCPreferenceVectorIndex<NumberVector>> indexfactory, double epsilon) Constructor.

Method Summary

Modifier and Type	Method and Description
Class<? super HiSC.HiSCClusterOrderEntry>	getEntryType() Get the entry type.
TypeInformation[]	getInputTypeRestriction() Get the input type restriction used for negotiating the data query.
protected Logging	getLogger() Get the (STATIC) logger for this class.
protected Collection<HiSC.HiSCClusterOrderEntry>	getNeighborsForDBID(Relation<V> relation, DBID id) Compute the neighbors for the given DBID.
protected HiSC.HiSCClusterOrderEntry	makeSeedEntry(Relation<V> relation, DBID objectID) Create the initial element to seed the algorithm.
ClusterOrderResult<HiSC.HiSCClusterOrderEntry>	run(Relation<V> relation) Run OPTICS on the database.
double	weightedDistance(V v1, V v2, long[] weightVector) Computes the weighted distance between the two specified vectors according to the given preference vector.

Methods inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.optics.GeneralizedOPTICS

expandClusterOrder, getMinPts

Methods inherited from class de.lmu.ifi.dbs.elki.algorithm.AbstractAlgorithm

makeParameterDistanceFunction, run

Methods inherited from class java.lang.Object

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

Methods inherited from interface de.lmu.ifi.dbs.elki.algorithm.clustering.optics.OPTICSTypeAlgorithm

run

Field Detail

LOG

private static final Logging LOG

The logger for this class.

indexfactory

private IndexFactory<V extends NumberVector,HiSCPreferenceVectorIndex<NumberVector>> indexfactory

Factory to produce

index

private HiSCPreferenceVectorIndex<NumberVector> index

Instantiated index.

relation

private Relation<V extends NumberVector> relation

Relation we are currently processing.

alpha

private double alpha

Holds the maximum diversion allowed.

Constructor Detail

HiSC

public HiSC(IndexFactory<V,HiSCPreferenceVectorIndex<NumberVector>> indexfactory,
            double epsilon)

Constructor.

Parameters:

indexfactory - HiSC index factory

Method Detail

run

public ClusterOrderResult<HiSC.HiSCClusterOrderEntry> run(Relation<V> relation)

Description copied from class: GeneralizedOPTICS

Run OPTICS on the database.

Overrides:

run in class GeneralizedOPTICS<V extends NumberVector,HiSC.HiSCClusterOrderEntry>

Parameters:

relation - Relation

Returns:

Result

makeSeedEntry

protected HiSC.HiSCClusterOrderEntry makeSeedEntry(Relation<V> relation,
                                                   DBID objectID)

Description copied from class: GeneralizedOPTICS

Create the initial element to seed the algorithm.

Specified by:

makeSeedEntry in class GeneralizedOPTICS<V extends NumberVector,HiSC.HiSCClusterOrderEntry>

Parameters:

relation - Data relation

objectID - Object ID

Returns:

Seed element.

getNeighborsForDBID

protected Collection<HiSC.HiSCClusterOrderEntry> getNeighborsForDBID(Relation<V> relation,
                                                                     DBID id)

Description copied from class: GeneralizedOPTICS

Compute the neighbors for the given DBID.

Specified by:

getNeighborsForDBID in class GeneralizedOPTICS<V extends NumberVector,HiSC.HiSCClusterOrderEntry>

Parameters:

relation - Data relation

id - Current object ID

Returns:

Neighbors

weightedDistance

public double weightedDistance(V v1,
                               V v2,
                               long[] weightVector)

Computes the weighted distance between the two specified vectors according to the given preference vector.

Parameters:

v1 - the first vector

v2 - the second vector

weightVector - the preference vector

Returns:

the weighted distance between the two specified vectors according to the given preference vector

getEntryType

public Class<? super HiSC.HiSCClusterOrderEntry> getEntryType()

Description copied from interface: OPTICSTypeAlgorithm

Get the entry type. Needed for type checking (i.e. is number distance)

Returns:

entry type

getInputTypeRestriction

public TypeInformation[] getInputTypeRestriction()

Description copied from class: AbstractAlgorithm

Get the input type restriction used for negotiating the data query.

Specified by:

getInputTypeRestriction in interface Algorithm

Specified by:

getInputTypeRestriction in class AbstractAlgorithm<ClusterOrderResult<HiSC.HiSCClusterOrderEntry>>

Returns:

Type restriction

getLogger

protected Logging getLogger()

Description copied from class: AbstractAlgorithm

Get the (STATIC) logger for this class.

Specified by:

getLogger in class AbstractAlgorithm<ClusterOrderResult<HiSC.HiSCClusterOrderEntry>>

Returns:

the static logger