Evolutionary clustering of relational data

Issue title: Advances in Intelligent Agent Systems

Guest editors: J.M. Benítezx, V. Loiay and F. Marcelloniz

Article type: Research Article

Authors: Horta, Danilo^{; *} | Campello, Ricardo J. G. B.

Affiliations: Department of Computer Sciences, University of São Paulo at São Carlos, SP 13560-970, Brazil | [x] Department of Computer Science and Artificial Intelligence, CITIC-UGR, Universidad de Granada, Granada, Spain | [y] Department of Mathematics and Computer Science, University of Salerno, Salerno, Italy | [z] Department of Information Engineering, University of Pisa, Pisa, Italy

Correspondence: [*] Corresponding author. Tel.: +55 (16) 3373 8161; E-mail: [email protected]

Abstract: This paper is concerned with the computational efficiency of clustering algorithms when the data set to be clustered is described by a proximity matrix only (relational data) and the number of clusters must be automatically estimated from such data. Two relational versions of an evolutionary algorithm for clustering are derived and compared against two systematic (pseudo-exhaustive) approaches that can also be used to automatically estimate the number of clusters in relational data. The computational complexities of the algorithms are discussed and an extensive collection of experiments involving 18 artificial and two real data sets is reported and analyzed.

Keywords: Evolutionary algorithms, relational data clustering

DOI: 10.3233/HIS-2010-0119

Journal: International Journal of Hybrid Intelligent Systems, vol. 7, no. 4, pp. 261-281, 2010