de.lmu.ifi.dbs.elki.math.dimensionsimilarity

Class HiCSDimensionSimilarity

java.lang.Object

de.lmu.ifi.dbs.elki.math.dimensionsimilarity.HiCSDimensionSimilarity

All Implemented Interfaces:

DimensionSimilarity<NumberVector<?>>, Parameterizable

public class HiCSDimensionSimilarity
extends Object
implements DimensionSimilarity<NumberVector<?>>

Use the statistical tests as used by HiCS to arrange dimensions.

Based on:
F. Keller, E. Müller, and K. Böhm.
HiCS: High Contrast Subspaces for Density-Based Outlier Ranking.
In ICDE, pages 1037–1048, 2012.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	HiCSDimensionSimilarity.Parameterizer Parameterization class.

Field Summary

Fields

Modifier and Type	Field and Description
private double	alpha Alpha threshold
private int	m Monte-Carlo iterations
private RandomFactory	rnd Random generator
private GoodnessOfFitTest	statTest Statistical test to use

Constructor Summary

Constructors

Constructor and Description
HiCSDimensionSimilarity(GoodnessOfFitTest statTest, int m, double alpha, RandomFactory rnd) Constructor.

Method Summary

Methods

Modifier and Type	Method and Description
private ArrayList<ArrayDBIDs>	buildOneDimIndexes(Relation<? extends NumberVector<?>> relation, DBIDs ids, DimensionSimilarityMatrix matrix) Calculates "index structures" for every attribute, i.e. sorts a ModifiableArray of every DBID in the database for every dimension and stores them in a list
private double	calculateContrast(Relation<? extends NumberVector<?>> relation, DBIDs subset, ArrayDBIDs subspaceIndex1, ArrayDBIDs subspaceIndex2, int dim1, int dim2, Random random) Calculates the actual contrast of a given subspace
void	computeDimensionSimilarites(Database database, Relation<? extends NumberVector<?>> relation, DBIDs subset, DimensionSimilarityMatrix matrix) Compute the dimension similarity matrix

Methods inherited from class java.lang.Object

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

Field Detail

m

private int m

Monte-Carlo iterations

alpha

private double alpha

Alpha threshold

statTest

private GoodnessOfFitTest statTest

Statistical test to use

rnd

private RandomFactory rnd

Random generator

Constructor Detail

HiCSDimensionSimilarity

public HiCSDimensionSimilarity(GoodnessOfFitTest statTest,
                       int m,
                       double alpha,
                       RandomFactory rnd)

Constructor.

Parameters:

statTest - Test function

m - Number of monte-carlo iterations

alpha - Alpha threshold

rnd - Random source

Method Detail

computeDimensionSimilarites

public void computeDimensionSimilarites(Database database,
                               Relation<? extends NumberVector<?>> relation,
                               DBIDs subset,
                               DimensionSimilarityMatrix matrix)

Description copied from interface: DimensionSimilarity

Compute the dimension similarity matrix

Specified by:

computeDimensionSimilarites in interface DimensionSimilarity<NumberVector<?>>

Parameters:

database - Database context

relation - Relation

subset - DBID subset (for sampling / selection)

matrix - Matrix to fill

buildOneDimIndexes

private ArrayList<ArrayDBIDs> buildOneDimIndexes(Relation<? extends NumberVector<?>> relation,
                                       DBIDs ids,
                                       DimensionSimilarityMatrix matrix)

Calculates "index structures" for every attribute, i.e. sorts a ModifiableArray of every DBID in the database for every dimension and stores them in a list

Parameters:

relation - Relation to index

ids - IDs to use

matrix - Matrix (for dimension subset)

Returns:

List of sorted objects

calculateContrast

private double calculateContrast(Relation<? extends NumberVector<?>> relation,
                       DBIDs subset,
                       ArrayDBIDs subspaceIndex1,
                       ArrayDBIDs subspaceIndex2,
                       int dim1,
                       int dim2,
                       Random random)

Calculates the actual contrast of a given subspace

Parameters:

relation - Data relation

subset - Subset to process

subspaceIndex1 - Index of first subspace

subspaceIndex2 - Index of second subspace

dim1 - First dimension

dim2 - Second dimension

random - Random generator

Returns:

Contrast