A Markov Chain Monte Carlo Procedure for Efficient Bayesian Inference on the Phase-Type Aging Model

Cong Nie, Xiaoming Liu, Serge Provost () and Jiandong Ren
Additional contact information
Cong Nie: Department of Statistical and Actuarial Sciences, The University of Western Ontario, London, ON N6A 3K7, Canada
Xiaoming Liu: Department of Statistical and Actuarial Sciences, The University of Western Ontario, London, ON N6A 3K7, Canada
Serge Provost: Department of Statistical and Actuarial Sciences, The University of Western Ontario, London, ON N6A 3K7, Canada
Jiandong Ren: Department of Statistical and Actuarial Sciences, The University of Western Ontario, London, ON N6A 3K7, Canada

Stats, 2025, vol. 8, issue 3, 1-22

Abstract: The phase-type aging model (PTAM) belongs to a class of Coxian-type Markovian models that can provide a quantitative description of well-known aging characteristics that are part of a genetically determined, progressive, and irreversible process. Due to its unique parameter structure, estimation via the MLE method presents a considerable estimability issue, whereby profile likelihood functions are flat and analytically intractable. In this study, a Markov chain Monte Carlo (MCMC)-based Bayesian methodology is proposed and applied to the PTAM, with a view to improving parameter estimability. The proposed method provides two methodological extensions based on an existing MCMC inference method. First, we propose a two-level MCMC sampling scheme that makes the method applicable to situations where the posterior distributions do not assume simple forms after data augmentation. Secondly, an existing data augmentation technique for Bayesian inference on continuous phase-type distributions is further developed in order to incorporate left-truncated data. While numerical results indicate that the proposed methodology improves parameter estimability via sound prior distributions, this approach may also be utilized as a stand-alone statistical model-fitting technique.

Keywords: parameter estimation; phase-type aging model; Markov chain Monte Carlo; Bayesian inference; data augmentation; Gibbs sampler; rejection sampling (search for similar items in EconPapers)
JEL-codes: C1 C10 C11 C14 C15 C16 (search for similar items in EconPapers)
Date: 2025
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2571-905X/8/3/77/pdf (application/pdf)
https://www.mdpi.com/2571-905X/8/3/77/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jstats:v:8:y:2025:i:3:p:77-:d:1733688

Access Statistics for this article

Stats is currently edited by Mrs. Minnie Li

More articles in Stats from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().