A0437
Title: Beta-binomial stick-breaking non-parametric prior
Authors: Ramses Mena - Universidad Nacional Autonoma De Mexico (Mexico) [presenting]
Abstract: A new class of nonparametric prior distributions, termed the Beta-Binomial stick-breaking process, is proposed. An appealing discrete random probability measure arises by allowing the underlying length random variables to be dependent through a Beta marginals Markov chain. The chain's dependence parameter controls the ordering of the stick-breaking weights and thus tunes the model's label-switching ability. Also, the resulting class contains the Dirichlet process and the Geometric process priors as particular cases, which is of interest for MCMC implementations by tuning this parameter. Some model properties are discussed, and a density estimation algorithm is proposed and tested with simulated datasets.
