On initializations for the Minkowski Weighted K-Means

Amorim, Renato Cordeiro de and Komisarczuk, Peter (2012) On initializations for the Minkowski Weighted K-Means. In: Advances in Intelligent Data Analysis XI - 11th International Symposium, IDA 2012, 25-27 Oct 2012, Helsinki, Finland.

Full text not available from this repository.


Minkowski Weighted K-Means is a variant of K-Means set in the Minkowski space, automatically computing weights for features at each cluster. As a variant of K-Means, its accuracy heavily depends on the initial centroids fed to it. In this paper we discuss our experiments comparing six initializations, random and five other initializations in the Minkowski space, in terms of their accuracy, processing time, and the recovery of the Minkowski exponent p.

We have found that the Ward method in the Minkowski space tends to outperform other initializations, with the exception of low-dimensional Gaussian Models with noise features. In these, a modified version of intelligent K-Means excels.

Item Type: Conference or Workshop Item (Paper)
ISSN: 0302-9743
ISBN: 9783642341557
Identifier: 10.1007/978-3-642-34156-4_6
Page Range: pp. 45-55
Identifier: 10.1007/978-3-642-34156-4_6
Keywords: Minkowski K-Means, K-Means Initializations, Lp Space, Minkowski Space, Feature Weighting, Noise Features, intelligent K-Means, Ward Method
Subjects: Computing
Depositing User: Vani Aul
Date Deposited: 21 Mar 2014 16:07
Last Modified: 28 Aug 2021 07:17
URI: https://repository.uwl.ac.uk/id/eprint/762

Actions (login required)

View Item View Item