This talk has captions. You can remove these by pressing CC on the video toolbar.
Name: Martina Favero
Talk Title: Asymptotic analysis of sampling probabilities and backward simulation algorithms for coalescent models
Abstract: Observing a certain configuration of genetic material in a large sample is an event with a small probability which can be estimated by importance sampling based on backward simulation of coalescent processes. In this talk, a novel framework for the analysis of asymptotic properties of these backward simulation algorithms is presented.
We start by showing that the sampling probabilities under the Kingman coalescent decay polynomially in the sample size. Then we present a weak convergence result for a sequence of Markov chains related to the coalescent, including associated cost chains. Finally we illustrate how these results can be applied to analyse asymptotic properties of backward sampling algorithms, in particular the asymptotic behaviour of importance sampling weights. This talk is based on joint work with H. Hult.
This talk is an invited talk at EVA 2021. View the programme here.