Packages

c

org.apache.spark.mllib.clustering

StreamingKMeansModel

class StreamingKMeansModel extends KMeansModel with Logging

StreamingKMeansModel extends MLlib's KMeansModel for streaming algorithms, so it can keep track of a continuously updated weight associated with each cluster, and also update the model by doing a single iteration of the standard k-means algorithm.

The update algorithm uses the "mini-batch" KMeans rule, generalized to incorporate forgetfulness (i.e. decay). The update rule (for each cluster) is:

$$ \begin{align} c_{t+1} &= [(c_t * n_t * a) + (x_t * m_t)] / [n_t + m_t] \\ n_{t+1} &= n_t * a + m_t \end{align} $$

Where c_t is the previously estimated centroid for that cluster, n_t is the number of points assigned to it thus far, x_t is the centroid estimated on the current batch, and m_t is the number of points assigned to that centroid in the current batch.

The decay factor 'a' scales the contribution of the clusters as estimated thus far, by applying a as a discount weighting on the current point when evaluating new incoming data. If a=1, all batches are weighted equally. If a=0, new centroids are determined entirely by recent data. Lower values correspond to more forgetting.

Decay can optionally be specified by a half life and associated time unit. The time unit can either be a batch of data or a single data point. Considering data arrived at time t, the half life h is defined such that at time t + h the discount applied to the data from t is 0.5. The definition remains the same whether the time unit is given as batches or points.

Annotations
@Since( "1.2.0" )
Source
StreamingKMeans.scala
Linear Supertypes
Logging, KMeansModel, PMMLExportable, Serializable, Serializable, Saveable, AnyRef, Any
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. StreamingKMeansModel
  2. Logging
  3. KMeansModel
  4. PMMLExportable
  5. Serializable
  6. Serializable
  7. Saveable
  8. AnyRef
  9. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. All

Instance Constructors

  1. new StreamingKMeansModel(clusterCenters: Array[Vector], clusterWeights: Array[Double])
    Annotations
    @Since( "1.2.0" )

Value Members

  1. val clusterCenters: Array[Vector]
    Definition Classes
    StreamingKMeansModelKMeansModel
    Annotations
    @Since( "1.2.0" )
  2. val clusterWeights: Array[Double]
    Annotations
    @Since( "1.2.0" )
  3. def computeCost(data: RDD[Vector]): Double

    Return the K-means cost (sum of squared distances of points to their nearest center) for this model on the given data.

    Return the K-means cost (sum of squared distances of points to their nearest center) for this model on the given data.

    Definition Classes
    KMeansModel
    Annotations
    @Since( "0.8.0" )
  4. val distanceMeasure: String
    Definition Classes
    KMeansModel
    Annotations
    @Since( "2.4.0" )
  5. def k: Int

    Total number of clusters.

    Total number of clusters.

    Definition Classes
    KMeansModel
    Annotations
    @Since( "0.8.0" )
  6. def predict(points: JavaRDD[Vector]): JavaRDD[Integer]

    Maps given points to their cluster indices.

    Maps given points to their cluster indices.

    Definition Classes
    KMeansModel
    Annotations
    @Since( "1.0.0" )
  7. def predict(points: RDD[Vector]): RDD[Int]

    Maps given points to their cluster indices.

    Maps given points to their cluster indices.

    Definition Classes
    KMeansModel
    Annotations
    @Since( "1.0.0" )
  8. def predict(point: Vector): Int

    Returns the cluster index that a given point belongs to.

    Returns the cluster index that a given point belongs to.

    Definition Classes
    KMeansModel
    Annotations
    @Since( "0.8.0" )
  9. def save(sc: SparkContext, path: String): Unit

    Save this model to the given path.

    Save this model to the given path.

    This saves:

    • human-readable (JSON) model metadata to path/metadata/
    • Parquet formatted data to path/data/

    The model may be loaded using Loader.load.

    sc

    Spark context used to save model data.

    path

    Path specifying the directory in which to save this model. If the directory already exists, this method throws an exception.

    Definition Classes
    KMeansModelSaveable
    Annotations
    @Since( "1.4.0" )
  10. def toPMML(): String

    Export the model to a String in PMML format

    Export the model to a String in PMML format

    Definition Classes
    PMMLExportable
    Annotations
    @Since( "1.4.0" )
  11. def toPMML(outputStream: OutputStream): Unit

    Export the model to the OutputStream in PMML format

    Export the model to the OutputStream in PMML format

    Definition Classes
    PMMLExportable
    Annotations
    @Since( "1.4.0" )
  12. def toPMML(sc: SparkContext, path: String): Unit

    Export the model to a directory on a distributed file system in PMML format

    Export the model to a directory on a distributed file system in PMML format

    Definition Classes
    PMMLExportable
    Annotations
    @Since( "1.4.0" )
  13. def toPMML(localPath: String): Unit

    Export the model to a local file in PMML format

    Export the model to a local file in PMML format

    Definition Classes
    PMMLExportable
    Annotations
    @Since( "1.4.0" )
  14. val trainingCost: Double
    Definition Classes
    KMeansModel
    Annotations
    @Since( "2.4.0" )
  15. def update(data: RDD[Vector], decayFactor: Double, timeUnit: String): StreamingKMeansModel

    Perform a k-means update on a batch of data.

    Perform a k-means update on a batch of data.

    Annotations
    @Since( "1.2.0" )