cbr.similarity
Class EuclideanDistanceMetric

java.lang.Object
  cbr.similarity.EuclideanDistanceMetric

All Implemented Interfaces:

SimilarityMetric

public class EuclideanDistanceMetric
extends Object
implements SimilarityMetric

Provides implementation for a Euclidean Distance Metric. The following formula is used:
Similarity = 1 - (|value1 - value2| / MAX_VALUE)
Which produces a normalised, inverted similarity value between two values.

Constructor Summary

EuclideanDistanceMetric(double max)
Creates a Euclidean Distance Metric given a value for the maximum distance between two values.

Method Summary

double	compare(Feature f1, Feature f2) Derives the normalised, inverted 1-dimensional Euclidean distance between two numerical feature values
double	getMax() Returns the maximum value, used in the equation for this metric.
void	setMax(double max) Specify the maximum difference between two numerical values

Methods inherited from class java.lang.Object

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

Constructor Detail

EuclideanDistanceMetric

public EuclideanDistanceMetric(double max)

Creates a Euclidean Distance Metric given a value for the maximum distance between two values.

Parameters:

max - the maximum value between two feature values

Method Detail

getMax

public double getMax()

Returns the maximum value, used in the equation for this metric.

Returns:

the maximum value

setMax

public void setMax(double max)

Specify the maximum difference between two numerical values

Parameters:

max - the maximum value

compare

public double compare(Feature f1,
                      Feature f2)

Derives the normalised, inverted 1-dimensional Euclidean distance between two numerical feature values

Specified by:

compare in interface SimilarityMetric

Parameters:

f1 - a case feature

f2 - another case feature

Returns:

1.0 is exactly similar, 0.0 is entirely dissimilar