基于误差指标的我国城镇化率最优组合预测  

Optimal Mix Forecasting of Urbanization Rate in China Based on Error Index

作　　者：马瑛琪 孙善辉[1] MA Ying - qi, SUN Shan- hui(Faculty of Mathematics and Statistics, Suzhou University, Suzhou 23400)

机构地区：[1]宿州学院数学与统计学院,安徽宿州234000

出　　处：《阴山学刊（自然科学版）》2018年第3期18-22,共5页Yinshan Academic Journal(Natural Science Edition)

基　　金：金融数学-安徽省高校大规模在线开放课程(MOOC)示范项目;质量工程项目(2016mooc301);安徽省2017年度高校优秀青年骨干人才国内访学研修项目(gxfx2017127)

摘　　要：以1978-2016年我国总人口和城镇人口数据为样本对我国城镇化率的变化趋势进行研究,并用指数平滑、多项式拟合、灰色预测三种方法对我国2017-2020年的城镇化率进行预测;进一步地构建基于误差指标的最优组合预测模型对同样时间段的城镇化率进行预测,模型检验表明,最优组合预测模型能拥有更小的预测误差,预测效果相比单项预测更加准确.从结果可以看出:2017-2020年我国城镇化率稳步逐年上升,到2020年我国人口城镇化率将达到62.97%,为政府科学决策提供依据.Based on the data of China＇s total population and urban population from 1978 to 2016,the trend of urbanization rate in China from 1978 to 2016 is studied and the urbanization of our country from 2017 to 2020 is predicted by exponential smoothing,polynomial fitting and gray forecasting. Further,this paper builds the optimal mix forecasting model based on the error index and forecast the urbanization rate in the same period of time. The results show that the forecasting result of the optimal mix forecasting model based on the error index is more accurate than the single forecasting method. With the continuous expansion of urbanization in our country,the rate of urbanization will steadily increase from 2017 to 2020,and it will reach 62. 97% by 2020,which provides the basis for the government＇s scientific decision-making.

关 键 词：城镇化率 指数平滑法 多项式拟合 灰色预测 组合预测 

分 类 号：O29[理学—应用数学] F201[理学—数学]