Bayesian mixture models in a longitudinal setting for analysing sheep CAT scan images

Alston, C.L.; Mengersen, Kerrie; Robert, Christian P.; Thompson, J.M.; Littlefield, P.J.; Perry, D.; Ball, A.J.

Alston, C.L.; Mengersen, Kerrie; Robert, Christian P.; Thompson, J.M.; Littlefield, P.J.; Perry, D.; Ball, A.J. (2007), Bayesian mixture models in a longitudinal setting for analysing sheep CAT scan images, Computational Statistics and Data Analysis, 51, 9, p. 4282–4296. http://dx.doi.org/10.1016/j.csda.2006.05.013

Type

Article accepté pour publication ou publié

Date

2007

Journal name

Computational Statistics and Data Analysis

Volume

Number

Publisher

Elsevier

Pages

4282–4296

Publication identifier

http://dx.doi.org/10.1016/j.csda.2006.05.013

Metadata

Show full item record

Author(s)

Alston, C.L.

Mengersen, Kerrie

Robert, Christian P.
CEntre de REcherches en MAthématiques de la DEcision [CEREMADE]
Thompson, J.M.

Littlefield, P.J.

Perry, D.

Ball, A.J.

Abstract (EN)

CAT scanning is used in longitudinal animal science experiments to assess possible changes to carcase composition induced by treatment over given time periods. A hierarchical Bayesian mixture model can be used to analyse the CAT scan data in terms of the proportion of each tissue type present in a scan. In this paper we present an extension to the hierarchical Bayesian mixture model in which estimated parameters from neighbouring CAT scans can be incorporated into the current model. These models are demonstrated using two examples.

Subjects / Keywords

Density estimation; Gibbs sampling; Markov Chain Monte Carlo; Markov random field; Metropolis–Hastings algorithm; Posterior simulation; Perturbation parameters