de.lmu.ifi.dbs.elki.math.linearalgebra.pca

Class PCAFilteredAutotuningRunner<V extends NumberVector<?>>

java.lang.Object

de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCARunner<V>

de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCAFilteredRunner<V>

de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCAFilteredAutotuningRunner<V>

Type Parameters:

V - vector type

All Implemented Interfaces:

InspectionUtilFrequentlyScanned, Parameterizable

@Reference(authors="H.-P. Kriegel, P. Kr\u00f6ger, E. Schubert, A. Zimek",
           title="A General Framework for Increasing the Robustness of PCA-based Correlation Clustering Algorithms",
           booktitle="Proceedings of the 20th International Conference on Scientific and Statistical Database Management (SSDBM), Hong Kong, China, 2008",
           url="http://dx.doi.org/10.1007/978-3-540-69497-7_27")
public class PCAFilteredAutotuningRunner<V extends NumberVector<?>>
extends PCAFilteredRunner<V>

Performs a self-tuning local PCA based on the covariance matrices of given objects. At most the closest 'k' points are used in the calculation and a weight function is applied. The number of points actually used depends on when the strong eigenvectors exhibit the clearest correlation.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
(package private) static class	PCAFilteredAutotuningRunner.Cand Candidate
static class	PCAFilteredAutotuningRunner.Parameterizer<V extends NumberVector<?>> Parameterization class.

Field Summary

Fields inherited from class de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCAFilteredRunner

BIG_ID, PCA_EIGENPAIR_FILTER, SMALL_ID

Fields inherited from class de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCARunner

covarianceMatrixBuilder, PCA_COVARIANCE_MATRIX

Constructor Summary

Constructors

Constructor and Description
PCAFilteredAutotuningRunner(CovarianceMatrixBuilder<V> covarianceMatrixBuilder, EigenPairFilter eigenPairFilter, double big, double small) Constructor.

Method Summary

Methods

Modifier and Type	Method and Description
private <D extends NumberDistance<D,?>> void	assertSortedByDistance(DistanceDBIDResult<D> results) Ensure that the results are sorted by distance.
private double	computeExplainedVariance(FilteredEigenPairs filteredEigenPairs) Compute the explained variance for a FilteredEigenPairs.
PCAFilteredResult	processIds(DBIDs ids, Relation<? extends V> database) Run PCA on a collection of database IDs.
<D extends NumberDistance<D,?>> PCAFilteredResult	processQueryResult(DistanceDBIDResult<D> results, Relation<? extends V> database) Run PCA on a QueryResult Collection.

Methods inherited from class de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCAFilteredRunner

getEigenPairFilter, processCovarMatrix, processEVD

Methods inherited from class de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCARunner

getCovarianceMatrixBuilder, processDatabase, setCovarianceMatrixBuilder

Methods inherited from class java.lang.Object

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

Constructor Detail

PCAFilteredAutotuningRunner

public PCAFilteredAutotuningRunner(CovarianceMatrixBuilder<V> covarianceMatrixBuilder,
                           EigenPairFilter eigenPairFilter,
                           double big,
                           double small)

Constructor.

Parameters:

covarianceMatrixBuilder - Covariance matrix builder

eigenPairFilter - Eigen pair filter

big - Replacement for large values

small - Replacement for small values

Method Detail

processIds

public PCAFilteredResult processIds(DBIDs ids,
                           Relation<? extends V> database)

Description copied from class: PCAFilteredRunner

Run PCA on a collection of database IDs.

Overrides:

processIds in class PCAFilteredRunner<V extends NumberVector<?>>

Parameters:

ids - a collection of ids

database - the database used

Returns:

PCA result

processQueryResult

public <D extends NumberDistance<D,?>> PCAFilteredResult processQueryResult(DistanceDBIDResult<D> results,
                                                                   Relation<? extends V> database)

Description copied from class: PCAFilteredRunner

Run PCA on a QueryResult Collection.

Overrides:

processQueryResult in class PCAFilteredRunner<V extends NumberVector<?>>

Type Parameters:

D - distance type

Parameters:

results - a collection of QueryResults

database - the database used

Returns:

PCA result

computeExplainedVariance

private double computeExplainedVariance(FilteredEigenPairs filteredEigenPairs)

Compute the explained variance for a FilteredEigenPairs.

Parameters:

filteredEigenPairs - Filtered eigenpairs

Returns:

explained variance by the strong eigenvectors.

assertSortedByDistance

private <D extends NumberDistance<D,?>> void assertSortedByDistance(DistanceDBIDResult<D> results)

Ensure that the results are sorted by distance.

Type Parameters:

D - distance type

Parameters:

results - Results to process