Research Repository

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

Cordeiro de Amorim, R and Makarenkov, V and Mirkin, B (2016) 'A-Wardpβ: Effective hierarchical clustering using the Minkowski metric and a fast k-means initialisation.' Information Sciences, 370-37. 343 - 354. ISSN 0020-0255

[img]
Preview
Text
1611.01060v1.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (1MB) | Preview

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 Ward p algorithms to the situation where the feature weight exponent can differ from the exponent of the Minkowski distance. This new method, called A-Ward pβ , 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 pβ provides better cluster recovery than both Ward and Ward p .

Item Type: Article
Uncontrolled Keywords: Initialisation algorithm Minkowski metric Hierarchical clustering Feature weighting
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Faculty of Science and Health > Computer Science and Electronic Engineering, School of
Depositing User: Elements
Date Deposited: 18 Sep 2017 11:40
Last Modified: 13 Dec 2017 14:46
URI: http://repository.essex.ac.uk/id/eprint/20363

Actions (login required)

View Item View Item