On a class of Bayesian nonparametric priors derived by subordination of Stable processes
Annalisa Cerquetti · 2009
We investigate a new class of Bayesian nonparametric priors derived by convolution mixture (composition) of the (positive) Stable r.v. by an inde-pendent ID r.v. belonging to the family of Generalized Gamma convolutions (Bondesson, 1992). We rely on the study proposed in James (2006) and on recent results for posterior analysis of BNP priors obtained by normalization (James et al., 2005, 2009). We derive posterior analysis, predictive distribution and the specific form of the induced exchangeable partition probability function.