it.uniroma2.sag.kelp.learningalgorithm.clustering.kmeans

Class LinearKMeansEngine

java.lang.Object

it.uniroma2.sag.kelp.learningalgorithm.clustering.kmeans.LinearKMeansEngine

All Implemented Interfaces:

ClusteringAlgorithm

public class LinearKMeansEngine
extends Object
implements ClusteringAlgorithm

Implements the K-means Clustering Algorithm, that works on an Explicit feature Space. It differs from the it.uniroma2.sag.kelp.learningalgorithm.clustering.kernelbasedkmeans.KernelBasedKMeansEngine as, at each iteration, it explicitly computes the centroid of each cluster. The computational cost is thus O(Ikn), where:
- I: is the number of iterations
- k: the number of desired clusters
- n: the number of input examples

while the complexity of it.uniroma2.sag.kelp.learningalgorithm.clustering.kernelbasedkmeans.KernelBasedKMeansEngine is O(n^2). For more details on the K-mean Algorithm, please refer to: https://en.wikipedia.org/wiki/K-means_clustering

Author:

Danilo Croce

Constructor Summary

Constructors

Constructor and Description
LinearKMeansEngine()
LinearKMeansEngine(String representationName, int k, int maxIterations)

Method Summary

Methods

Modifier and Type	Method and Description
ClusterList	cluster(Dataset dataset) It starts the clustering process exploiting the provided dataset
ClusterList	cluster(Dataset dataset, ExampleSelector seedSelector) It starts the clustering process exploiting the provided dataset
int	getK()
int	getMaxIterations()
String	getRepresentationName()
void	setK(int k)
void	setMaxIterations(int maxIterations)
void	setRepresentationName(String representationName)

Methods inherited from class java.lang.Object

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

Constructor Detail

LinearKMeansEngine

public LinearKMeansEngine()

LinearKMeansEngine

public LinearKMeansEngine(String representationName,
                  int k,
                  int maxIterations)

Parameters:

representationName - The representation name containing the vector used by the algorithm

k - The number of expected clusters

maxIterations - The maximum number of iterations

Method Detail

cluster

public ClusterList cluster(Dataset dataset)

Description copied from interface: ClusteringAlgorithm

It starts the clustering process exploiting the provided dataset

Specified by:

cluster in interface ClusteringAlgorithm

Returns:

the data instances grouped in clusters

cluster

public ClusterList cluster(Dataset dataset,
                  ExampleSelector seedSelector)

Description copied from interface: ClusteringAlgorithm

It starts the clustering process exploiting the provided dataset

Specified by:

cluster in interface ClusteringAlgorithm

seedSelector - the seed selector

Returns:

the data instances grouped in clusters

getK

public int getK()

getMaxIterations

public int getMaxIterations()

getRepresentationName

public String getRepresentationName()

setK

public void setK(int k)

setMaxIterations

public void setMaxIterations(int maxIterations)

setRepresentationName

public void setRepresentationName(String representationName)