Probabilistic clustering of interval data

Article type: Research Article

Authors: Brito, Paula^{a; *} | Silva, A. Pedro Duarte^b | Dias, José G.^c

Affiliations: [a] FEP & LIAAD INESC TEC, Universidade do Porto, Portugal | [b] School of Economics and Management and CEGE, Universidade Católica Portuguesa, Portugal | [c] Instituto Universitário de Lisboa (ISCTE-IUL), BRU, Portugal

Correspondence: [*] Corresponding author: Paula Brito, Fac. de Economia do Porto, Rua Dr. Roberto Frias, 4200-464 Porto, Portugal. Fax: +351 5505050; E-mail: [email protected].

Abstract: In this paper we address the problem of clustering interval data, adopting a model-based approach. To this purpose, parametric models for interval-valued variables are used which consider configurations for the variance-covariance matrix that take the nature of the interval data directly into account. Results, both on synthetic and empirical data, clearly show the well-founding of the proposed approach. The method succeeds in finding parsimonious heterocedastic models which is a critical feature in many applications. Furthermore, the analysis of the different data sets made clear the need to explicitly consider the intrinsic variability present in interval data.

Keywords: Clustering methods, finite mixture models, interval-valued variable, intrinsic variability, symbolic data

DOI: 10.3233/IDA-150718

Journal: Intelligent Data Analysis, vol. 19, no. 2, pp. 293-313, 2015