Open Access
Translator Disclaimer
2021 On a Dirichlet Process Mixture Representation of Phase-Type Distributions
Daniel Ayala, Leonardo Jofré, Luis Gutiérrez, Ramsés H. Mena
Author Affiliations +
Bayesian Anal. Advance Publication 1-26 (2021). DOI: 10.1214/21-BA1272

Abstract

An explicit representation of phase-type distributions as an infinite mixture of Erlang distributions is introduced. The representation unveils a novel and useful connection between a class of Bayesian nonparametric mixture models and phase-type distributions. In particular, this sheds some light on two hot topics, estimation techniques for phase-type distributions, and the availability of closed-form expressions for some functionals related to Dirichlet process mixture models. The power of this connection is illustrated via a posterior inference algorithm to estimate phase-type distributions, avoiding some difficulties with the simulation of latent Markov jump processes, commonly encountered in phase-type Bayesian inference. On the other hand, closed-form expressions for functionals of Dirichlet process mixture models are illustrated with density and renewal function estimation, related to the optimal salmon weight distribution of an aquaculture study.

Funding Statement

The work of the first author was supported by “Becas Doctorado Nacional CONICYT 2017 Folio No. 21171601”. The work of the third author was supported by “Proyecto REDES ETAPA INICIAL Convocatoria 2017 REDI170094” and by ANID–Millennium Science Initiative Program–NCN17_059. The fourth author acknowledges the support of CONTEX project 2018-9B.

Acknowledgments

The authors are grateful for the valuable comments made by two anonymous referees, the Associate Editor and the Editor. The authors thank professor Ricardo Olea Ortega by providing the salmon weights data set.

Citation

Download Citation

Daniel Ayala. Leonardo Jofré. Luis Gutiérrez. Ramsés H. Mena. "On a Dirichlet Process Mixture Representation of Phase-Type Distributions." Bayesian Anal. Advance Publication 1 - 26, 2021. https://doi.org/10.1214/21-BA1272

Information

Published: 2021
First available in Project Euclid: 8 June 2021

Digital Object Identifier: 10.1214/21-BA1272

Subjects:
Primary: 62G05 , 62M05
Secondary: 60J28

Keywords: Bayesian nonparametrics , Erlang distribution , mixture model , Renewal function

JOURNAL ARTICLE
26 PAGES


SHARE
Advance Publication
Back to Top