Title: A Bayesian semi-parametric approach to estimate elliptical distributions
Authors: J Miguel Marin - University Carlos III (Spain)
Pepa Ramirez Cobo - Universidad de Cadiz (Spain)
Fabrizio Ruggeri - CNR - IMATI (Italy)
M Remedios Sillero-Denamiel - University of Seville (Spain) [presenting]
Michael Wiper - Universidad Carlos III de Madrid (Spain)
Abstract: Elliptical distributions are generalisations of the multivariate normal distribution to both longer tails and elliptical contours. We examine how we can use Bayesian methods to perform semi-parametric inference for elliptical distributions. We show how our approach might be used to test whether data are generated from an elliptical distribution or to test whether elliptical data could come from a particular elliptical distribution such as the normal or Student's t.