Packages

c

org.apache.spark.mllib.clustering

BisectingKMeans

class BisectingKMeans extends Logging

A bisecting k-means algorithm based on the paper "A comparison of document clustering techniques" by Steinbach, Karypis, and Kumar, with modification to fit Spark. The algorithm starts from a single cluster that contains all points. Iteratively it finds divisible clusters on the bottom level and bisects each of them using k-means, until there are k leaf clusters in total or no leaf clusters are divisible. The bisecting steps of clusters on the same level are grouped together to increase parallelism. If bisecting all divisible clusters on the bottom level would result more than k leaf clusters, larger clusters get higher priority.

Annotations
@Since( "1.6.0" )
Source
BisectingKMeans.scala
See also

Steinbach, Karypis, and Kumar, A comparison of document clustering techniques, KDD Workshop on Text Mining, 2000.

Linear Supertypes
Logging, AnyRef, Any
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. BisectingKMeans
  2. Logging
  3. AnyRef
  4. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. All

Instance Constructors

  1. new BisectingKMeans()

    Constructs with the default configuration

    Constructs with the default configuration

    Annotations
    @Since( "1.6.0" )

Value Members

  1. def getDistanceMeasure: String

    The distance suite used by the algorithm.

    The distance suite used by the algorithm.

    Annotations
    @Since( "2.4.0" )
  2. def getK: Int

    Gets the desired number of leaf clusters.

    Gets the desired number of leaf clusters.

    Annotations
    @Since( "1.6.0" )
  3. def getMaxIterations: Int

    Gets the max number of k-means iterations to split clusters.

    Gets the max number of k-means iterations to split clusters.

    Annotations
    @Since( "1.6.0" )
  4. def getMinDivisibleClusterSize: Double

    Gets the minimum number of points (if greater than or equal to 1.0) or the minimum proportion of points (if less than 1.0) of a divisible cluster.

    Gets the minimum number of points (if greater than or equal to 1.0) or the minimum proportion of points (if less than 1.0) of a divisible cluster.

    Annotations
    @Since( "1.6.0" )
  5. def getSeed: Long

    Gets the random seed.

    Gets the random seed.

    Annotations
    @Since( "1.6.0" )
  6. def run(data: JavaRDD[Vector]): BisectingKMeansModel

    Java-friendly version of run().

  7. def run(input: RDD[Vector]): BisectingKMeansModel

    Runs the bisecting k-means algorithm.

    Runs the bisecting k-means algorithm.

    input

    RDD of vectors

    returns

    model for the bisecting kmeans

    Annotations
    @Since( "1.6.0" )
  8. def setDistanceMeasure(distanceMeasure: String): BisectingKMeans.this.type

    Set the distance suite used by the algorithm.

    Set the distance suite used by the algorithm.

    Annotations
    @Since( "2.4.0" )
  9. def setK(k: Int): BisectingKMeans.this.type

    Sets the desired number of leaf clusters (default: 4).

    Sets the desired number of leaf clusters (default: 4). The actual number could be smaller if there are no divisible leaf clusters.

    Annotations
    @Since( "1.6.0" )
  10. def setMaxIterations(maxIterations: Int): BisectingKMeans.this.type

    Sets the max number of k-means iterations to split clusters (default: 20).

    Sets the max number of k-means iterations to split clusters (default: 20).

    Annotations
    @Since( "1.6.0" )
  11. def setMinDivisibleClusterSize(minDivisibleClusterSize: Double): BisectingKMeans.this.type

    Sets the minimum number of points (if greater than or equal to 1.0) or the minimum proportion of points (if less than 1.0) of a divisible cluster (default: 1).

    Sets the minimum number of points (if greater than or equal to 1.0) or the minimum proportion of points (if less than 1.0) of a divisible cluster (default: 1).

    Annotations
    @Since( "1.6.0" )
  12. def setSeed(seed: Long): BisectingKMeans.this.type

    Sets the random seed (default: hash value of the class name).

    Sets the random seed (default: hash value of the class name).

    Annotations
    @Since( "1.6.0" )