Do We Have Signals? Revealing Substantial Cohort Change in Mortality Modelling
Conference
64th ISI World Statistics Congress - Ottawa, Canada
Format: CPS Abstract
Keywords: machine learning, modelling, mortality, multivariate control chart
Session: CPS 77 - Statistics and mortality
Wednesday 19 July 8:30 a.m. - 9:40 a.m. (Canada/Eastern)
Abstract
Mortality modelling is a practical method for the government and various fields such as economics, actuarial, and official statistics. Through this modelling, we can get a picture of mortality up to age-specific for a particular year. However, some information on the phenomenon may remain in the residuals vector and unrevealed from the models. We handle this issue by employing a multivariate control chart to discover substantial cohort changes in mortality behaviour that the models still need to collect. The Hotelling T2 control chart is applied to the externally studentised deviance model, which is already optimised using a machine-learning decision tree. This study shows a mortality model with the lowest MSE, MAPE and Deviance by accomplishing simulations in various countries. In addition, the model is more sensitive in detecting signals on the control chart so that we can perform a decomposition to determine the attribute of death in the specific age group in a particular year. The case study in the decomposition uses data from the country of Indonesia. The overall results demonstrate that our method of processing and producing mortality models with machine learning can be a solution for developing countries or countries with limited mortality data to produce accurate predictions through monitoring control charts.