Translated Poisson approximation to equilibrium distributions of Markov population processes (Q607605)

The authors consider the behavior of equilibrium distributions of density dependent Markov population processes in which the transition rates depend on the density of individuals. The Kolmogorov distance between the equilibrium distribution is known [\textit{A. D. Barbour}, Adv. Appl. Probab. 12, 591--614 (1980; Zbl 0434.60084)] to have a bound of order \(O(1/\sqrt{n})\). The authors prove that this normal approximation can be substantially strengthened. The line of reasoning is based on the Stein-Chen method and Dynkin's formula and leads to the approximating distribution in the form of a translated Poisson distribution with the same variance and almost the same mean. In spite of much stronger metric and weaker assumptions the error bounds are of the same order. The results are illustrated by a simple example in which an immigration birth and death process with birth occurring in groups is considered.