org.deeplearning4j.clustering

Class KMeansClustering

java.lang.Object

org.deeplearning4j.clustering.KMeansClustering

All Implemented Interfaces:

Serializable

public class KMeansClustering
extends Object
implements Serializable

Shamelessly based on: https://github.com/pmerienne/trident-ml/blob/master/src/main/java/com/github/pmerienne/trident/ml/clustering/KMeans.java adapted to jblas double matrices

Author:

Adam Gibson

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
KMeansClustering(Integer nbCluster)
KMeansClustering(Integer nbCluster, Class<? extends DistanceFunction> clazz)

Method Summary

Methods

Modifier and Type	Method and Description
Integer	classify(org.jblas.DoubleMatrix features)
protected org.jblas.DoubleMatrix	computeDxs() For each features in KMeans#initFeatures, compute D(x), the distance between x and the nearest center that has already been chosen.
org.jblas.DoubleMatrix	distribution(org.jblas.DoubleMatrix features)
org.jblas.DoubleMatrix	getCentroids()
protected void	initCentroids() Init clusters using the k-means++ algorithm.
protected void	initIfPossible(org.jblas.DoubleMatrix features)
protected boolean	isReady()
protected Integer	nearestCentroid(org.jblas.DoubleMatrix features)
void	reset()
Integer	update(org.jblas.DoubleMatrix features)

Methods inherited from class java.lang.Object

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

Constructor Detail

KMeansClustering

public KMeansClustering(Integer nbCluster,
                Class<? extends DistanceFunction> clazz)

KMeansClustering

public KMeansClustering(Integer nbCluster)

Method Detail

classify

public Integer classify(org.jblas.DoubleMatrix features)

update

public Integer update(org.jblas.DoubleMatrix features)

distribution

public org.jblas.DoubleMatrix distribution(org.jblas.DoubleMatrix features)

getCentroids

public org.jblas.DoubleMatrix getCentroids()

nearestCentroid

protected Integer nearestCentroid(org.jblas.DoubleMatrix features)

isReady

protected boolean isReady()

initIfPossible

protected void initIfPossible(org.jblas.DoubleMatrix features)

initCentroids

protected void initCentroids()

Init clusters using the k-means++ algorithm. (Arthur, D. and Vassilvitskii, S. (2007). "k-means++: the advantages of careful seeding".

computeDxs

protected org.jblas.DoubleMatrix computeDxs()

For each features in KMeans#initFeatures, compute D(x), the distance between x and the nearest center that has already been chosen.

Returns:

reset

public void reset()