On Seeking Consensus Between Document Similarity Measures

Article type: Research Article

Authors: Kłopotek, Mieczysław^{; *}

Affiliations: Institute of Computer Science, Polish Academy of Sciences, ul. Jana Kazimierza 5, 01-248 Warszawa, Poland. [email protected]

Correspondence: [*] Address for correspondece: Institute of Computer Science, Polish Academy of Sciences, ul. Jana Kazimierza 5, 01-248 Warszawa, Poland.

Abstract: This paper investigates the application of consensus clustering and meta-clustering to the set of all possible partitions of a data set. We show that when using a “complement” of Rand Index as a measure of cluster similarity, the total-separation partition, putting each element in a separate set, is chosen.

Keywords: cluster analysis, partitioning, clustering, consensus functions, ensemble, knowledge reuse, unsupervised learning, meta-clustering

DOI: 10.3233/FI-2017-1597

Journal: Fundamenta Informaticae, vol. 156, no. 1, pp. 43-68, 2017