Name: Gabriel Núñez Antonio
Institution: UAM-I, Mexico
E-mail: gab.nuneza@gmail.com
Co-authors: Emiliano Geneyro Squarzon
Abstract:
This article introduces a new Bayesian nonparametric model for directional data based on Dirichlet process (DP) mixtures of gamma distributions. That model is a projected DP mixture of multivariate gamma distributions and from a methodological point of view it is important to describe directional data restricted to the first orthant of unit sphere. We show how to carry out inference for the model based on a slice sampling scheme. Our proposal is illustrated using simulated data sets.