Online Program

Return to main conference page
Friday, February 21
Fri, Feb 21, 5:15 PM - 6:30 PM
Regency EF
Poster Session 2 and Refreshments

Modeling and Forecasting Mortality Rates Data from Upper-Middle-Income Economies: A Machine Learning Approach (304046)

View Presentation View Presentation

*Ahmad Talafha, Western Michigan University 
*Emmanuel Thompson, Southeast Missouri State University 

Keywords: Lee-Carter, Machine learning, Stochastic mortality models, Upper middle-income economies

Since Gompertz’s 1825 publication on law of mortality, several stochastic mortality models have been proposed for a wide range of application from demographic predictions to financial management. In recent years, computers have altered work in nearly every sector of the economy. Many fields are now incorporating machine learning techniques in their work and the results have been very promising. Available research in mortality modeling using machine learning techniques have originated predominantly from high income economies, not much has been done for populations within other income economies. Using mortality data from upper middle-income economies, this study seeks to contribute to filling the gap by using machine learning algorithms to fit and study the improvement in the goodness of fit of two standard stochastic mortality models: Lee-Carter and Renshaw-Haberman. The mean squared error and the root mean squared logarithm error are used as metrics to assess the forecasting potential of the fitted models.