Can Tree-Based Models Improve GAPC Mortality Models’ Forecasting Accuracy?

Generalized Age-Period-Cohort (GAPC) models are mortality models incorporates stochasticity which can be represented in generalized linear or non-linear context. By fitting the data to either mortality model, one can make forecasts for the future under the extrapolation framework. Previous research indicates that the machine learning techniques can be applied to improve forecasting ability of such mortality models using different training/testing time periods. However, there is no consensus about generalizing this phenomenon on the improvement of fitted/forecasted mortality rates without depending on particular mortality model or models’ training/testing period. The idea in this study is to re-estimate the mortality rates obtained from each mortality models by applying tree-based machine learning methods within a procedure that creates suitable environment for integrating machine learning models for each GAPC models. This study shows that if the proper procedure is applied, the forecasting ability of each mortality model can be improved.