edu.ucla.sspace.clustering

Class BisectingKMeans

java.lang.Object

edu.ucla.sspace.clustering.BisectingKMeans

All Implemented Interfaces:

Clustering

public class BisectingKMeans
extends Object
implements Clustering

An implementation of the Bisecting K-Means algorithm, also known as Repeated Bisections. This implementation is based on the following paper:

Michael Steinbach, George Karypis, Vipin Kumar. "A comparison of document clustering techniques," in KDD Workshop on Text Mining, 200 This clustering algorithm improves upon the standard K-Means algorithm by taking a data set and repeatedly splitting the data points into two regions. Initially all data points are separated into two clusters. Then, until the desired number of clusters are created, the largest cluster is divided using K-Means with K equal to 2. This implementation relies on the KMeansClustering implementation. Any properties passed to this clustering method are passed onto the KMeansClustering algorithm, allowing the user to set the desired seeding method.

Author:

Keith Stevens

See Also:

KMeansClustering