Single and multi-population mortality models for Dutch data
In this thesis we investigate mortality models for both single and multiple populations, using either frequentist or Bayesian inference. We develop a Bayesian single-population mortality model for Dutch mortality data, and generalize this afterwards to a Bayesian two-population model.This Bayesian two-population model allows joint modelling for male and female mortality, or portfolio-specic mortality together with population data. The use of Bayesian inference has multipleadvantages in mortality modeling. For example, incorporating parameter uncertainty (cf. Solvency II) is straightforward and interactions between two populations are taken into account.The contribution of this thesis lies in explaining the technicalities of MCMC sampling for Bayesian single and multi-population mortality models, and applying these models to Dutch mortality data.