Application Lee-Carter Model to Chinese Population Data under Uncertainty of Model's Parameters

Lee-Carter model and its extended models are popular and important models in mortality researches, which use time series methods to predict mortality in the future. The predict results often including an interval with a probability to cover the mortality. However the estimated parameters in models are often considered certainty; in fact the estimated model's parameters have uncertainty. Many recent researches concern the uncertainty in mortality studies. In Chinese population mortality researches, uncertainty of model parameters is scarce. In this paper, through Lee-Carter model and Bootstrap, mortality and other correlatives are to be studied under uncertainty of model's parameters based on Chinese population data.