Abstract

TOVAR CUEVAS, José Rafael; PORTILLA YELA, Jennyfer  and  ACHCAR, Jorge Alberto. A Method to Select Bivariate Copula Functions. Rev.Colomb.Estad. [online]. 2019, vol.42, n.1, pp.61-80.  Epub May 23, 2019. ISSN 0120-1751.  https://doi.org/10.15446/rce.v42n1.71780.

Copula functions have been extensively used in applied statistics, becoming a good alternative for modeling the dependence of multivariate data. Each copula function has a different dependence structure. An important issue in these applications is the choice of an appropriate copula function model for each case where standard classical or Bayesian discrimination methods could be not appropriate to decide by the best copula. Considering only the special case of bivariate data, we propose a procedure obtained from a recently dependence measure introduced in the literature to select an appropriate copula for the statistical data analyses.

Keywords : Copula functions; Discrimination of copulas; Dependence measure; Ledwina measure; Selection method.

        · abstract in Spanish     · text in English     · English ( pdf )