基于IOWA算子的中国人口出生率的组合预测研究

The Research on Combination Forecast of Fertility Rate of China Based on IOWA Operator

根据1995年—2015年我国人口出生率数据,采用多元线性回归、指数平滑、ARMA模型3种预测方法,分别得出我国人口出生率2016年—2020年的单项预测值。引入诱导有序加权算术平均(IOWA)算子,基于误差平方和最小的组合预测模型,并且通过模型评价发现,该组合预测模型的各种误差指标小于单项预测值,因此,验证了模型的有效性。利用所建立的模型预测未来5年我国人口出生率。由于人口出生率关系到国家长期的经济可持续发展,通过预测结果可为政府提出相应建议。

This paper uses the multiple linear regression,exponential smoothing and ARMA model to forecast the single prediction value of fertility rate in China on 2016—2020 from a series of fertility rate data in China from 1995 to 2015.In this paper,the induced ordered weighted averaging(IOWA)operator has been introduced to establish the error square sum and the minimum combination forecasting model.From the model evaluation,it is found that the square error sum of the combined forecasting model is smaller than the single prediction value.Therefore,the validity of the model is verified.Using the established model,it can predict the fertility rate in next five years in China.Since one countrys fertility rate related to its longterm sustainable development of economy,the results from the forecast can help the government to propose some appropriate advice.