modeling and forecasting mortality rate with cohort effect andoverdispersion

In this paper, we present an extension of the poisson lee-carter modelfor predicting mortality rates that includes the cohort effect and an overdispersioncomponent. to compute the estimation of the model parameters and mortality ratepredictions, we use bayesian analysis and the no-u-turn sampler (nuts), which is anextension of hamiltonian monte carlo (hmc). we compare the prediction accuracyof the proposed model and the poisson log-normal lee-carter (plnlc) model usingreal mortality data from several european countries. we conclude that the proposedmodel dominates the plnlc model.