Forecasting and Analysis Mortality using Lee-Carter Model with application

We demonstrate here a model proposed by Lee and Carter (1992) for fit and to forecast mortality rates. This approach is used widely in demographical applications and academic literature because the structure of the Lee-Carter model allows for the construction of confidence intervals related to mortality and age-specific death protections. To improve the performance of the Lee-Carter model, several extensions to the original version have been proposed. In this paper, we use real data of mortality rates by gender in Kirkuk City in Iraq during the period 2006-2015 to apply a modification of the Lee-Carter model, which accommodates variations in age-specific parameters using a singular values decomposition method to estimate the Lee-Carter parameters model. We also use the autoregressive moving average (ARIMA) with the special case of Random walk with drift (RWD) model to forecast the general index for the time period 2015-2020. The paper further predicts the survival expectancy at birth for each gender. Our results found this survival expectancy to be increasing for age group (0-1) year and to be decreasing for age group (75-80) years. For the long-term forecast it is necessary for the field of Demography to obtain such predications, which depend however on the available data.