org.openimaj.math.statistics.distribution.metrics

Class SampledMultivariateDistanceComparator

java.lang.Object

org.openimaj.math.statistics.distribution.metrics.SampledMultivariateDistanceComparator

All Implemented Interfaces:

DistanceComparator<MultivariateDistribution>

public class SampledMultivariateDistanceComparator
extends Object
implements DistanceComparator<MultivariateDistribution>

By sampling a distribution and calculating the log liklihood of those samples against another distribution, construct a distance metric. This function uses MultivariateDistribution.estimateLogProbability(double[][])

Author:

Sina Samangooei (ss@ecs.soton.ac.uk)

Constructor Summary

Constructors

Constructor and Description
SampledMultivariateDistanceComparator()
SampledMultivariateDistanceComparator(int nsamples)
SampledMultivariateDistanceComparator(long seed, int nsamples)

Method Summary

Modifier and Type	Method and Description
double	compare(MultivariateDistribution o1, MultivariateDistribution o2) Compare two objects, returning a score or distance.
boolean	isDistance()

Methods inherited from class java.lang.Object

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

Constructor Detail

SampledMultivariateDistanceComparator

public SampledMultivariateDistanceComparator()

SampledMultivariateDistanceComparator

public SampledMultivariateDistanceComparator(int nsamples)

Parameters:

nsamples -

SampledMultivariateDistanceComparator

public SampledMultivariateDistanceComparator(long seed,
                                             int nsamples)

Parameters:

seed -

nsamples -

Method Detail

compare

public double compare(MultivariateDistribution o1,
                      MultivariateDistribution o2)

Description copied from interface: DistanceComparator

Compare two objects, returning a score or distance.

Specified by:

compare in interface DistanceComparator<MultivariateDistribution>

Parameters:

o1 - the first object

o2 - the second object

Returns:

a score or distance

isDistance

public boolean isDistance()

Specified by:

isDistance in interface DistanceComparator<MultivariateDistribution>

Returns:

true if the comparison is a distance; false if similarity.