Modeling And Prediction Of Population Life

Life expectancy per capita is one of the important indicators of the level of national economic development.As life expectancy increases,the ensuing problem is the aging of the population.For example,For example,the statutory retirement age of men in our country is 65 and the retirement age of women is 60.;Under the same legal working age,the longer life expectancy per capita means that the pressure of social support is greater.If you can understand in advance the trend of life expectancy per capita social pension plan will play a decisive role.In order to solve the prediction of life expectancy,it is of crucial importance to study the mortality rate of each age part of the elderly population.In order to accurately predict life expectancy,two modeling methods are used to model the mortality rate of the population,which are ARIMA model and Holt-Winters index model respectively;As the mortality rates for each age segment are usually variable,but the closer the age,the closer the natural mortality rate,so for the ARIMA model of the original data,let's first take the cosine similarity as a measure of age-group clustering,we get three clustering results,then select one age from each cluster as a representative data series of mortality modeling;finally,the life expectancy per capita is obtained through the transformation relationship between life expectancy and mortality.For the Holt-Winters model,the data is segmented first,then the periodic data sequence is constructed by the segmented data,and the periodic data sequence segment is taken as the Holt-Winters modeling data.Through the reasonable analysis and processing of the data,the ARIMA model was obtained by using the data from 1963 to 2015.The Holt-Winters model was obtained by straightening the data from 1983 to 2015.The relative error and absolute error of the mortality rate are used as evaluation indexes of the model's merits.Finally,the feasibility and validity of the ARIMA model are verified by the prediction of relative error and absolute error.The comparison between the modeling step and experimental results shows that the comprehensive performance of the Holt-Winters model is significantly better than that of the ARIMA model.The ARIMA model and Holt-Winters model modeling method can be effectively used in the modeling of population mortality.