Dynamic Forecasting the Mortality of China's Population on Quantile Auto Regression
2019.06.06Ming Zhao, Xiaojun Wang
【Abstract】
This paper improves the forecasting method of Lee-Carter model based on the data of the year 1994-2012from the National Bureau of Statistics. We build a Chinese mortality model with quantile auto regression method to forecast the mortality trend in the future. Comparing the predicted results with the mean regression, we can get more information to overcome the mortality underestimate and make the future mortality prediction more reasonable and reliable.
【Keywords】
Multivariate generalized Poisson, moment estimation, maximum likelihoodestimation, hypothesis test.