Document Type

Conference Proceeding

Publication Date

2006

Abstract

The unsupervised nature of cluster analysis means that objects can be clustered in many ways, allowing different clustering algorithms to generate vastly different results. To address this, clustering comparison methods have traditionally been used to quantify the degree of similarity between alternative clusterings. However, existing techniques utilize only the point memberships to calculate the similarity, which can lead to unintuitive results. They also cannot be applied to analyze clusterings which only partially share points, which can be the case in stream clustering. In this paper we introduce a new measure named ADCO, which takes into account density profiles for each attribute and aims to address these problems. We provide experiments to demonstrate this new measure can often provide a more reasonable similarity comparison between different clusterings than existing methods.

Comments

This paper was presented at the 19th Australian Joint Conference on Artificial Intelligence, Hobart, Australia, December 4-8, 2006.

The featured PDF document is the unpublished, peer-reviewed version of this article.

The featured abstract was published in the final version of this article, which appeared in Lecture Notes in Computer Science, volume 4304, pp. 342-351 and may be found at http://link.springer.com/chapter/10.1007%2F11941439_38 .

DOI

10.1007/11941439_38

Download


Share

COinS