Name: Jose Rafael Tovar
Institution: Universidad del Valle, Colombia
Co-authors: Llerzy Esneider Torres

One difficulty that arise when making inferences from the Bayesian approach is the obtaining of the value of the hyperparameters that completely define the probability distribution called a priori. The difficulty can be greater when the generating process that is studied is of continuous nature and the parameter of interest represents a dependency structure. A method is proposed to obtain values of the hyperparameters of the a priori distribution that models the natural behavior of the copula-type dependency parameter. To illustrate the proposed method, the problem of estimating the Gumbel-Barnett copula type dependence parameter was considered assuming three a priori distributions (Beta, Kumaraswamy, Truncated Gamma), 12 simulation scenarios were formed using three dependency levels and four sample sizes. The results obtained with the proposed method were compared with those generated using other strategies reported in the literature.