Q-Markov random probability measures and their posterior distributions
In this paper, we use the Markov property introduced in Balan and Ivanoff (J. Theoret. Probab. 15 (2002) 515) for set-indexed processes and we prove that a Markov prior distribution leads to a Markov posterior distribution. In particular, by proving that a neutral to the right prior distribution leads to a neutral to the right posterior distribution, we extend a fundamental result of Doksum (Ann. Probab. 2 (1974) 183) to arbitrary sample spaces.
Year of publication: |
2004
|
---|---|
Authors: | Balan, R. M. |
Published in: |
Stochastic Processes and their Applications. - Elsevier, ISSN 0304-4149. - Vol. 109.2004, 2, p. 295-316
|
Publisher: |
Elsevier |
Keywords: | Random probability measure Q-Markov process Transition system Dirichlet process Neutral to the right process |
Saved in:
Saved in favorites
Similar items by person
-
Asymptotic Results with Generalized Estimating Equations for Longitudinal data II
Balan, R. M., (2004)
-
Set-indexed processes with independent increments
Balan, R. M., (2002)
-
A strong Markov property for set-indexed processes
Balan, R. M., (2001)
- More ...