A-Ward: Effective hierarchical clustering using the Minkowski metric and a fast k-means initialisation

Renato Cordeiro De Amorim, Vladimir Makarenkov, Boris Mirkin

    Research output: Contribution to journalArticlepeer-review

    13 Citations (Scopus)
    16 Downloads (Pure)

    Abstract

    In this paper we make two novel contributions to hierarchical clustering. First, we introduce an anomalous pattern initialisation method for hierarchical clustering algorithms, called A-Ward, capable of substantially reducing the time they take to converge. This method generates an initial partition with a sufficiently large number of clusters. This allows the cluster merging process to start from this partition rather than from a trivial partition composed solely of singletons. Our second contribution is an extension of the Ward and Wardp algorithms to the situation where the feature weight exponent can differ from the exponent of the Minkowski distance. This new method, called A-Ward, is able to generate a much wider variety of clustering solutions. We also demonstrate that its parameters can be estimated reasonably well by using a cluster validity index. We perform numerous experiments using data sets with two types of noise, insertion of noise features and blurring within-cluster values of some features. These experiments allow us to conclude: (i) our anomalous pattern initialisation method does indeed reduce the time a hierarchical clustering algorithm takes to complete, without negatively impacting its cluster recovery ability; (ii) A-Ward provides better cluster recovery than both Ward and Wardp.

    Original languageEnglish
    Pages (from-to)343-354
    Number of pages12
    JournalInformation Sciences
    Volume370-371
    Early online date1 Aug 2016
    DOIs
    Publication statusPublished - 20 Nov 2016

    Keywords

    • Feature weighting
    • Hierarchical clustering
    • Initialisation algorithm
    • Minkowski metric

    Fingerprint

    Dive into the research topics of 'A-Ward: Effective hierarchical clustering using the Minkowski metric and a fast k-means initialisation'. Together they form a unique fingerprint.

    Cite this