Class AnnulusKMeans<V extends NumberVector>

  • Type Parameters:
    V - vector datatype
    All Implemented Interfaces:
    Algorithm, ClusteringAlgorithm<Clustering<KMeansModel>>, KMeans<V,​KMeansModel>

    @Reference(authors="J. Drake",title="Faster k-means clustering",booktitle="Faster k-means clustering",url="http://hdl.handle.net/2104/8826",bibkey="mathesis/Drake13") @Reference(authors="G. Hamerly and J. Drake",title="Accelerating Lloyd\u2019s Algorithm for k-Means Clustering",booktitle="Partitional Clustering Algorithms",url="https://doi.org/10.1007/978-3-319-09259-1_2",bibkey="doi:10.1007/978-3-319-09259-1_2")
    public class AnnulusKMeans<V extends NumberVector>
    extends HamerlyKMeans<V>
    Annulus k-means algorithm. A variant of Hamerly with an additional bound, based on comparing the norm of the mean and the norm of the points.

    This implementation could be further improved by precomputing and storing the norms of all points (at the cost of O(n) memory additionally).

    Reference:

    J. Drake
    Faster k-means clustering
    Masters Thesis

    G. Hamerly and J. Drake
    Accelerating Lloyd’s Algorithm for k-Means Clustering
    Partitional Clustering Algorithms

    Since:
    0.7.5
    Author:
    Erich Schubert