gov.sandia.cognition.learning.data.feature

Class MultivariateDecorrelator.DiagonalCovarianceLearner

java.lang.Object

gov.sandia.cognition.util.AbstractCloneableSerializable

gov.sandia.cognition.learning.data.feature.MultivariateDecorrelator.DiagonalCovarianceLearner

All Implemented Interfaces:

BatchLearner<java.util.Collection<? extends Vectorizable>,MultivariateDecorrelator>, CloneableSerializable, java.io.Serializable, java.lang.Cloneable

Enclosing class:

MultivariateDecorrelator

public static class MultivariateDecorrelator.DiagonalCovarianceLearner
extends AbstractCloneableSerializable
implements BatchLearner<java.util.Collection<? extends Vectorizable>,MultivariateDecorrelator>

The DiagonalCovarianceLearner class implements a BatchLearner object for a MultivariateDecorrelator.

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
static double	DEFAULT_DEFAULT_COVARIANCE The default value for default covariance is 1.0E-5.
protected double	defaultCovariance The default covariance.

Constructor Summary

Constructors

Constructor and Description
DiagonalCovarianceLearner() Creates a new MultivariateDecorrelator.DiagonalCovarianceLearner with the default value for default covariance.
DiagonalCovarianceLearner(double defaultCovariance) Creates a new MultivariateDecorrelator.DiagonalCovarianceLearner with the given value for default covariance.

Method Summary

Modifier and Type	Method and Description
double	getDefaultCovariance() Gets the default covariance value.
MultivariateDecorrelator	learn(java.util.Collection<? extends Vectorizable> values) Learns a MultivariateDecorrelator from the given values by computing the mean and variance for each dimension separately.
void	setDefaultCovariance(double defaultCovariance) Sets the default covariance value.

Methods inherited from class gov.sandia.cognition.util.AbstractCloneableSerializable

clone

Methods inherited from class java.lang.Object

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

Methods inherited from interface gov.sandia.cognition.util.CloneableSerializable

clone

Field Detail

DEFAULT_DEFAULT_COVARIANCE

public static final double DEFAULT_DEFAULT_COVARIANCE

The default value for default covariance is 1.0E-5.

See Also:

Constant Field Values

defaultCovariance

protected double defaultCovariance

The default covariance. Added to the diagonal to prevent it from becoming singular.

Constructor Detail

DiagonalCovarianceLearner

public DiagonalCovarianceLearner()

Creates a new MultivariateDecorrelator.DiagonalCovarianceLearner with the default value for default covariance.

DiagonalCovarianceLearner

public DiagonalCovarianceLearner(double defaultCovariance)

Creates a new MultivariateDecorrelator.DiagonalCovarianceLearner with the given value for default covariance.

Parameters:

defaultCovariance - The default covariance value. Added to the diagonal to prevent it from becoming singular

Method Detail

learn

public MultivariateDecorrelator learn(java.util.Collection<? extends Vectorizable> values)

Learns a MultivariateDecorrelator from the given values by computing the mean and variance for each dimension separately.

Specified by:

learn in interface BatchLearner<java.util.Collection<? extends Vectorizable>,MultivariateDecorrelator>

Parameters:

values - The values to use.

Returns:

The MultivariateDecorrelator computed from the given values with a diagonal covariance matrix.

getDefaultCovariance

public double getDefaultCovariance()

Gets the default covariance value. It is added to the diagonal terms of the covariance matrix to attempt to prevent them from becoming singular.

Returns:

The default covariance value.

setDefaultCovariance

public void setDefaultCovariance(double defaultCovariance)

Sets the default covariance value. It is added to the diagonal terms of the covariance matrix to attempt to prevent them from becoming singular.

Parameters:

defaultCovariance - The default covariance value. Must be non-negative.