基于邻近度的安徽省人均GDP组合预测模型
|
摘要
针对安徽省人均GDP预测问题,以安徽省2000—2018年人均GDP数据为研究区间,其中2000—2017年数据作为训练集,2018年数据作为测试集,提出了一类新的预测评价指标-邻近度及基于邻近度的组合预测模型,并引入一种新的组合权系数求解方法;首先对训练集进行单项预测,即对训练集数据进行指数预测、抛物线预测和移动平均预测,接下来对各单项预测值综合考虑,建立基于邻近度的加权几何平均组合预测模型,通过求解模型得出各单项预测权系数进而求出基于邻近度的组合预测值,最后分别在测试集和训练集上与其他预测方法预测结果进行比较,并预测安徽省2019—2021年人均GDP数据。
In order to solve the problem of per capita GDP forecasting in Anhui province,taking the per capita GDP data of Anhui province from 2000 to 2018 as the research sample,where the data from 2000 to 2017 are used as the training set and the data in 2018 as the testing set,a new forecasting evaluation index--adjacent degree and combination forecasting model based on adjacent degree was proposed,and a new combination weights coefficient solution method was introduced. First,the data is used to do three kinds of single forecasting,which is the exponential forecasting,parabolic forecasting and moving average forecasting. Next,three kinds of single forecasting models are considered comprehensively, constructing a weighted geometric average combination forecasting model based on adjacent degrees. The combination forecasting model is solved,then the weights of every single forecasting are obtained,and the combination forecasting value based on adjacent degree is received.Finally,comparing with other forecasting models in training set and testing set,2019—2021 per capita GDP data in Anhui province is forecasted.
In order to solve the problem of per capita GDP forecasting in Anhui province,taking the per capita GDP data of Anhui province from 2000 to 2018 as the research sample,where the data from 2000 to 2017 are used as the training set and the data in 2018 as the testing set,a new forecasting evaluation index--adjacent degree and combination forecasting model based on adjacent degree was proposed,and a new combination weights coefficient solution method was introduced. First,the data is used to do three kinds of single forecasting,which is the exponential forecasting,parabolic forecasting and moving average forecasting. Next,three kinds of single forecasting models are considered comprehensively, constructing a weighted geometric average combination forecasting model based on adjacent degrees. The combination forecasting model is solved,then the weights of every single forecasting are obtained,and the combination forecasting value based on adjacent degree is received.Finally,comparing with other forecasting models in training set and testing set,2019—2021 per capita GDP data in Anhui province is forecasted.
引文
[1] BATES J M. GRANGER C W J. Combination of Forecasts[J]. Operations Research Quarterly,1969,20(4):451—468
[2]唐小我.组合预测计算方法研究[J].预测,1991(4):35—39TANG X W. Research on Computational Methods of Combination Forecasting[J]. Forecasting,1991(4):35—39(in Chinese)
[3]陈华友.基于预测有效度的组合预测模型研究[J].预测,2001,20(3):72—73CHEN H Y. Research on Combination Forecasting Model Based on Effective Measure of Forecasting Methods[J].Forecasting,1991(3):35—39(in Chinese)
[4]谢开贵,周家启.组合预测模型的回归分析方法[J].统计研究,2003,26(1):62—65XIE K G,ZHOU J Q. Combination Forecasting Models Using Regression Analysis Method[J]. Statistical Research,2003,26(1):62—65(in Chinese)
[5] YAO W X,LI L H. A New Regression Model:Modal Linear Regression[J]. Scandinavian Journal of Statistics,2013,1—29
[6] YAO W X. A Note on EM Algorithm for Mixture Models[J]. Statistics and Probability Letters,2013, 83:519—526
[7]陈华友,朱家明,丁珍妮.组合预测模型和方法研究述[J].大学数学,2017,33(4):1—10CHEN H Y, ZHU J M, DING Z N, A Survey of Researches on Combination Forecasting Models and Methodolog[J]. College Mathematics,2017,33(4):1—10(in Chinese)
[8]詹英.组合预测方法在我国人均GDP中的应用[D].武汉:华中师范大学,2014ZHAN Y. Combination Forecasting Method in China’s Per Capita GDP Forecast[D]. Wuhan:Central China Normal University,2014(in Chinese)
[9]陈华友.基于L1范数的加权几何平均组合预测模型[J].安徽大学学报(自然科学版),2004,28(4):5—10CHEN H Y. Weighted Geometric Means Combination Forecasting Method Based on L1 Norm[J]. Journal of Anhui University(Natural Science Edition). 2004,28(4):5—10(in Chinese)
[2]唐小我.组合预测计算方法研究[J].预测,1991(4):35—39TANG X W. Research on Computational Methods of Combination Forecasting[J]. Forecasting,1991(4):35—39(in Chinese)
[3]陈华友.基于预测有效度的组合预测模型研究[J].预测,2001,20(3):72—73CHEN H Y. Research on Combination Forecasting Model Based on Effective Measure of Forecasting Methods[J].Forecasting,1991(3):35—39(in Chinese)
[4]谢开贵,周家启.组合预测模型的回归分析方法[J].统计研究,2003,26(1):62—65XIE K G,ZHOU J Q. Combination Forecasting Models Using Regression Analysis Method[J]. Statistical Research,2003,26(1):62—65(in Chinese)
[5] YAO W X,LI L H. A New Regression Model:Modal Linear Regression[J]. Scandinavian Journal of Statistics,2013,1—29
[6] YAO W X. A Note on EM Algorithm for Mixture Models[J]. Statistics and Probability Letters,2013, 83:519—526
[7]陈华友,朱家明,丁珍妮.组合预测模型和方法研究述[J].大学数学,2017,33(4):1—10CHEN H Y, ZHU J M, DING Z N, A Survey of Researches on Combination Forecasting Models and Methodolog[J]. College Mathematics,2017,33(4):1—10(in Chinese)
[8]詹英.组合预测方法在我国人均GDP中的应用[D].武汉:华中师范大学,2014ZHAN Y. Combination Forecasting Method in China’s Per Capita GDP Forecast[D]. Wuhan:Central China Normal University,2014(in Chinese)
[9]陈华友.基于L1范数的加权几何平均组合预测模型[J].安徽大学学报(自然科学版),2004,28(4):5—10CHEN H Y. Weighted Geometric Means Combination Forecasting Method Based on L1 Norm[J]. Journal of Anhui University(Natural Science Edition). 2004,28(4):5—10(in Chinese)
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |