Abstract
We present new methodology for drawing samples from a posterior distribution when the likelihood function is only specified up to a normalising constant. Our method is "on-line" as compared with alternative approaches to the problem which require "off-line" computations. Since it is needed to simulate from the "unknown distribution", perfect simulation algorithms become useful. We illustrate the method in cases whre the likelihood is given by a Markov point process model. Particularly, we consider semi-parametric Bayesian inference in connection to both inhomogeneous Markov point process models and pairwise interaction point processes.
Originalsprog | Engelsk |
---|---|
Titel | Proceedings from the International Conference on Spatial Point Process Modelling and Its Applications |
Redaktører | Baddeley, Gregori, Mateu, Stoica, Stoyan |
Forlag | Publicacions de la Universitat Jaume I |
Publikationsdato | 2004 |
Sider | 7-15 |
ISBN (Trykt) | 8480214759 |
Status | Udgivet - 2004 |
Begivenhed | International Conference on Spatial Point Process Modelling and Its Applications - Varighed: 19 maj 2010 → … |
Konference
Konference | International Conference on Spatial Point Process Modelling and Its Applications |
---|---|
Periode | 19/05/2010 → … |