Investigating predictive probabilities of Gibbs-type priors

1 MISTIS [2016-2019] - Modelling and Inference of Complex and Structured Stochastic Systems [2016-2019]
Inria Grenoble - Rhône-Alpes, LJK [2016-2019] - Laboratoire Jean Kuntzmann [2016-2019], Grenoble INP [2007-2019] - Institut polytechnique de Grenoble - Grenoble Institute of Technology [2007-2019]
Abstract : Gibbs-type priors are arguably the most 'natural' generalization of the Dirichlet prior. Among them the two parameter Poisson-Dirichlet prior certainly stands out for the simplicity and intuitiveness of its predictive probabilities. Given an observable sample of size $n$, in this paper we show that the predictive probabilities of any Gibbs-type prior admit a large $n$ approximation, with an error term vanishing as $o(1/n)$, which maintains the same mathematical tractability and interpretability as the predictive probabilities of the two parameter Poisson-Dirichlet prior. We discuss the use of our approximate predictive probabilities in connection with some recent work on Bayesian nonparametric inference for discovery probabilities.
Document type :
Conference papers
Domain :
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01667765
Contributor : Julyan Arbel <>
Submitted on : Tuesday, December 19, 2017 - 3:37:10 PM
Last modification on : Thursday, August 6, 2020 - 3:16:55 AM

Identifiers

• HAL Id : hal-01667765, version 1

Citation

Julyan Arbel. Investigating predictive probabilities of Gibbs-type priors. Mathematical Methods of Modern Statistics, Jul 2017, Marseille, France. ⟨hal-01667765⟩

Record views