The spine decomposition of branching processes and their applications

Rongli LIU,Yan-Xia REN
2018-01-01
Abstract:Due to their close relation to both biology and other branches of mathematics, branching processes, branching Markov processes and superprocesses, have been receiving more and more attention. Early studies of these processes mainly used analytic methods, which are not very transparent. More intuitive probabilistic arguments would be very helpful. In recent years, many people started to use probabilistic methods to study these processes. One of these probabilistic methods is the so-called spine method. In the supercritical cases, the spine is the trajectory of an immortal moving particle and the spine decomposition theorem says that, after a martingale change of measure, the transformed process can be decomposed in law as an immigration process along this spine. The spine method mainly uses the spine decomposition of the processes to reduce the study of a random number of sample paths of these processes to the study of one sample path. In this paper, we give a survey of the spine method and its applications to the three types of processes above.
What problem does this paper attempt to address?