textminer.clustering
Class kMeans

java.lang.Object
  |
  +--textminer.clustering.AbstractClusterer
        |
        +--textminer.clustering.kMeans

All Implemented Interfaces:

InterOutcomes

public final class kMeans

extends AbstractClusterer

The kMeans class is an implementation of k-means clustering.

The k-means clustering is an algorithm for partitioning (or clustering) N data points into k disjoint subsets s_{j} containing N_{j} data points so as to minimize the sum-of-squares error. It is comprised of a simple re-estimation procedure as follows:

• First, the data points are assigned at random to the k sets.
• Second, the centroid is computed for each set.

These two steps are alternated until a stopping criterion is met (i.e. when there is no further change in the assignment of the data points.)
For more information, see:

• [Bishop, 1995] Bishop, C. M., 1995. Neural Networks for Pattern Recognition, Oxford University Press.
• [Jain and Dubes, 1988] Jain, A.K., and Dubes, R.C.,1988, Algorithms for Clustering Data, Prentice Hall: New Jersey.

Since:

0.1

Version:

TextMiner 1.1

Author:

Young-Woo Seo (ywseo@cs.cmu.edu)

Field Summary

Fields inherited from class textminer.clustering.AbstractClusterer

data_repository, dataset_dir, docvec_repository, num_instances, result_repository, task_alias, task_name, vectorindex

Fields inherited from interface textminer.core.InterOutcomes

ext_condensed_index_file, ext_corpus_stat_file, ext_dvec_file, ext_fsmethod_file, ext_index_file, ext_judgment_file, ext_lexicon_file, ext_matrix_file, ext_model_file, ext_output_file, ext_result_file, ext_termdic_file, ext_vec_index_file

Constructor Summary

kMeans(SubtaskClustering subtasks)
Constructor of kMeans

Method Summary

void	doClustering() Perform kMeans
void	init(int k_size, int size_of_vector, int max_iterations) Initialize
void	make_results(java.lang.String filename) Write the clustering resutls the specified filename
void	print_clusters() Print the clustering results

Methods inherited from class textminer.clustering.AbstractClusterer

init_clusterer, log

Methods inherited from class java.lang.Object

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

Constructor Detail

kMeans

public kMeans(SubtaskClustering subtasks)

Constructor of kMeans

Parameters:

subtasks - specification for kMeans clustering

Method Detail

init

public void init(int k_size,
                 int size_of_vector,
                 int max_iterations)

Initialize

Parameters:

k_size - derised number of clusters

size_of_vector - number of components in a document vector

max_iterations - maximum number of iterations

doClustering

public void doClustering()

Perform kMeans

Specified by:

doClustering in class AbstractClusterer

print_clusters

public void print_clusters()

Print the clustering results

make_results

public void make_results(java.lang.String filename)

Write the clustering resutls the specified filename