Sankhya: The Indian Journal of Statistics

1996, Volume 58, Series B, Pt. 2, pp. 363--378

BAYESIAN ANALYSIS OF INVERSE GAUSSIAN NON-LINEAR REGRESSION BY SIMULATION

By

S. K. UPADHYAY, R. AGRAWAL, Benaras Hindu University, Varanasi

and

A. F. M. SMITH, Imperial College, London

SUMMARY.  This paper examines and illustrates two simulation methods, namely the Gibbs sampler  and sampling importance-resampling, for the Bayesian analysis of an inverse Gaussian nonlinear regression model commonly arising in accelerated life testing experiments. The approaches are conceptually simple and have wide applicability from the perspective of reliability practitioners.

AMS (1991) subject classification.  62F15, 62J02.

Key words and phrases. Accelerated testing, Gibbs sampler, sampling-importance-resampling, inverse Gaussian non-linear regression.

Full paper (PDF)

This article in Mathematical Reviews.