On Initializations for the Minkowski Weighted K-Means

• Published in: Advances in Intelligent Data Analysis XI pp. 45 - 55
• Main Authors: de Amorim, Renato Cordeiro, Komisarczuk, Peter
• Format: Book Chapter
• Language: English
• Published: Berlin, Heidelberg Springer Berlin Heidelberg 2012
• Series: Lecture Notes in Computer Science
• Subjects: Feature Weighting; intelligent K-Means; K-Means Initializations; Minkowski K-Means; Minkowski Space; Noise Features; Space; Ward Method
• ISBN: 9783642341557; 3642341551
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-642-34156-4_6

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.