RT Generic
T1 A semi-parametric model for circular data based on mixtures of beta distributions
A1 Carnicero, José Antonio
A1 Wiper, Michael Peter
A2 Universidad Carlos III de Madrid. Departamento de Estadística, 
AB This paper introduces a new, semi-parametric model for circular data, based on mixtures ofshifted, scaled, beta (SSB) densities. This model is more general than the Bernstein polynomialdensity model which is well known to provide good approximations to any density with finitesupport and it is shown that, as for the Bernstein polynomial model, the trigonometric moments ofthe SSB mixture model can all be derived.Two methods of fitting the SSB mixture model are considered. Firstly, a classical, maximumlikelihood approach for fitting mixtures of a given number of SSB components is introduced. TheBayesian information criterion is then used for model selection. Secondly, a Bayesian approachusing Gibbs sampling is considered. In this case, the number of mixture components is selectedvia an appropriate deviance information criterion.Both approaches are illustrated with real data sets and the results are compared with thoseobtained using Bernstein polynomials and mixtures of von Mises distributions.
YR 2008
FD 2008-03
LK https://hdl.handle.net/10016/2362
UL https://hdl.handle.net/10016/2362
LA eng
DS e-Archivo
RD 27 jul. 2024