TIME SERIES MODELING AND PREDICTION OF LIFE EXPECTANCY RATE AT BIRTH IN PAKISTAN

Author(s) : MUHAMMAD WASIM AMIR, ALI HASSAN MALIK , ZEESHAN RAZA , MUHAMMAD SAAD KHAN

Abstract:

The aim of this study attempted to shed light on the issues such as forecasting of life expectancy rate
at birth in Pakistan. Data on life expectancy rate at birth are collected over 37 years from period 1980 to 2017.
A variety of time series models are applied to find the most appropriate model for forecasting the life expectancy
rate at birth (LEB) of Pakistan. In the study, the autoregressive integrated moving average ARIMA (3, 2, 2) is
found to be the most adequate model for forecasting the LEB of Pakistan. The best model is selected based on
the various model selection’s tools. Furthermore, different test of runs and Jarque-Bera (JB) tests are used to
justify the assumption of the randomness and normality of residuals. Based on the best ARIMA (3, 2, 2) model
the forecasted value of the LEB in 2020,2022, and 2025 are 67.05,67.70, and 68.97 respectively.