Abstract: In this paper, the APC (age period cohort) model with year of birth effect is used to fit and predict China's population mortality data. Firstly, the parameters of APC model are estimated by maximum likelihood method. Through residual analysis of the model, it is found that the model captures the effects of age, calendar year and birth year, and the data fitting effect is good. Then the ARIMA time series model is used to fit the time index and the birth year index of the mortality model. The optimal ARIMA model is selected by AICc value for prediction, and then the predicted value of mortality rate is obtained, and the prediction effect is compared according to the real value of mortality. Finally, the life expectancy of population is predicted by using the predicted mortality value, and part of the age life expectancy table is drawn. The results show that: from the perspective of fitting effect and prediction effect, APC model is suitable for predicting China's population mortality rate. At the same time, we find that China's population life expectancy tends to be gradually extended, and the aging problem is becoming increasingly serious, which will bring huge financial pressure to China's life insurance enterprises and pension institutions.Abstract: In this paper, the APC (age period cohort) model with year of birth effect is used to fit and predict China's population mortality data. Firstly, the parameters of APC model are estimated by maximum likelihood method. Through residual analysis of the model, it is found that the model captures the effects of age, calendar year and birth year, and th...Show More
