A Fuzzy Variant of the Rand Index for Comparing Clustering Structures

Eyke Hullermeier (1) , Maria Rifqi (2)

1 Computational Intelligence Group
2 MALIRE - Machine Learning and Information Retrieval

In this paper, we introduce a fuzzy extension of the Rand index, a well-known measure for comparing two clustering structures. In contrast to an existing proposal, which is restricted to the comparison of a fuzzy partition with a non-fuzzy reference partition, our extension is able to compare two proper fuzzy partitions with each other. Elaborating on the formal properties of our fuzzy Rand index, we show that it exhibits desirable metrical properties.

Domaines: Intelligence artificielle [cs.AI]

Eyke Hullermeier, Maria Rifqi. A Fuzzy Variant of the Rand Index for Comparing Clustering Structures. Joint 2009 International Fuzzy Systems Association World Congress and 2009 European Society of Fuzzy Logic and Technology Conference, IFSA-EUSFLAT 2009, Jul 2009, Lisbon, Portugal. pp.1294-1298. ⟨hal-01072704⟩