中国人口预测的具有外生变量的半参数回归模型  

A Semi-parametric Autoregression Model with an Exogenous Variable for Population of China

作　　者：韩玉涛[1] 杨万才[1] 武新乾[1] 

机构地区：[1]河南科技大学数学与统计学院,河南洛阳471003

出　　处：《数理统计与管理》2012年第5期791-798,共8页Journal of Applied Statistics and Management

基　　金：河南省科技发展规划项目(092300410178);河南省教育厅自然科学研究计划项目(2008A110005;2010B110009);河南科技大学博士科研启动基金项目(09001322)

摘　　要：本文提出建立中国人口预测的具有外生变量的半参数回归模型,基于线性回归理论选取的滞后显著性变量为线性部分,外生变量做为非参数部分,利用多项式样条估计得到了半参数回归方程,对中国1972-2000年人口进行拟合比较,并且对中国2001-2009年人口分三种情况进行了预测对比,结果表明基于中国1952-2005年人口建立的半参数模型拟合和预测的精度均较高,最后对中国2010-2015年的人口进行了预测。A semi-parametric autoregression model with an exogenous variable for population of China is presented in this paper. The lag significant variables selected by linear regression theories are the linear part of the model, and the exogenous variable is the nonparametric component. A corresponding semiparametric autoregression equation is obtained through polynomial spline estimation. The population of China from 1972 to 2000 is fitted, whereafter the observations in 2001-2009 are predicted and analysised under three cases. Simulated results show that the precision of fit and forecast for the established model based on the population of 1952-2005 are better than that of the other models. Furthemore, the total population of China from 2010 to 2015 is predicted.

关 键 词：半参数回归 外生变量 多项式样条估计 人口 预测 

分 类 号：O212.7[理学—概率论与数理统计]