Examples of weka.clusterers.FarthestFirst

weka.clusterers.FarthestFirst

Cluster data using the FarthestFirst algorithm.

For more information see:

Hochbaum, Shmoys (1985). A best possible heuristic for the k-center problem. Mathematics of Operations Research. 10(2):180-184.

Sanjoy Dasgupta: Performance Guarantees for Hierarchical Clustering. In: 15th Annual Conference on Computational Learning Theory, 351-363, 2002.

Notes:
- works as a fast simple approximate clusterer
- modelled after SimpleKMeans, might be a useful initializer for it

BibTeX:

 @article{Hochbaum1985, author = {Hochbaum and Shmoys}, journal = {Mathematics of Operations Research}, number = {2}, pages = {180-184}, title = {A best possible heuristic for the k-center problem}, volume = {10}, year = {1985} } @inproceedings{Dasgupta2002, author = {Sanjoy Dasgupta}, booktitle = {15th Annual Conference on Computational Learning Theory}, pages = {351-363}, publisher = {Springer}, title = {Performance Guarantees for Hierarchical Clustering}, year = {2002} }

Valid options are:

 -N <num> number of clusters. (default = 2).

 -S <num> Random number seed. (default 1)

@author Bernhard Pfahringer (bernhard@cs.waikato.ac.nz) @version $Revision: 5538 $ @see RandomizableClusterer

  public ClusteringResult performClustering(Instances dataset,
      ParameterSet parameters) {


    List<Integer> clusters = new ArrayList<Integer>();
    String[] options = new String[2];
    FarthestFirst clusterer = new FarthestFirst();


    int numberOfGroups = parameters.getParameter(
        FarthestFirstClustererParameters.numberOfGroups).getValue();
    options[0] = "-N";
    options[1] = String.valueOf(numberOfGroups);


    try {
      clusterer.setOptions(options);
      clusterer.buildClusterer(dataset);
      Enumeration e = dataset.enumerateInstances();
      while (e.hasMoreElements()) {
        clusters.add(clusterer.clusterInstance((Instance) e
            .nextElement()));
      }
      ClusteringResult result = new ClusteringResult(clusters, null,
          clusterer.numberOfClusters(), parameters.getParameter(
              EMClustererParameters.visualization).getValue());
      return result;
    } catch (Exception ex) {
      logger.log(Level.SEVERE, null, ex);
      return null;

Examples of weka.clusterers.FarthestFirst

Related Classes of weka.clusterers.FarthestFirst