Bivariate power series distributions Bayesian estimation Trimmed likelihood Multitype branching processes
Issue Date:
2015
Publisher:
Institute of Mathematics and Informatics at the Bulgarian Academy of Sciences
Citation:
Pliska Studia Mathematica Bulgarica, Vol. 24, No 1, (2015), 73p-88p
Abstract:
We consider two-type branching stochastic processes with offspring distributions from the bivariate poisson distribution family. We study the parametric estimation under different sampling schemes – when the entire family tree is observed and when observations only on the generation sizes are made. We use a randomized algorithm to switch from the generation sizes scheme
to family tree observations and apply it in the context of the Bayesian approach. The considered estimation methods are illustrated via a simulational and computational example. 2010 Mathematics Subject Classification: Primary 60J80; Secondary 62F15, 62F35.
