“十二五”时期河南省经济发展预测研究

发布时间：2018-05-12 22:34

本文选题：线性自回归 + 半参数自回归　；参考：《河南科技大学》2013年硕士论文

【摘要】：合理预测“十二五”时期河南省经济发展趋势和状况，为省委、省政府和有关部门进行科学决策和有效管理提供定量依据，对促进中原经济区建设具有重要的现实意义。半参数回归模型既含有参数分量又含有非参数分量，它不但保留了参数模型易于解释的优点，，而且还具有建模的灵活性。本文基于半参数回归理论与方法建立了河南省国内生产总值(GDP)指数、人均GDP指数及第一、二、三产业指数(1952年为100)的半参数自回归模型，并且对相关经济指标进行了预测分析，具体内容如下： 1．详细给出了河南省GDP指数的线性自回归建模过程，以线性自回归模型的显著变量为半参数模型的线性主部变量、非显著变量为非参数函数变量，采用多项式样条估计方法建立了GDP指数的半参数趋势自回归模型。该模型对1985-2010年GDP指数拟合的均方标准误差(RMSE)为61.45，平均绝对百分误差(MAPE)为1.14%，预测结果表明：“十二五”时期，河南省GDP指数年均增长速度为9.26%，仍然呈现着较为平稳的发展态势。 2．类似于GDP指数的建模方法，建立了人均GDP指数的半参数自回归模型，其中滞后一、三、四、六阶变量为线性主部变量，滞后五阶变量为非参数函数变量。该模型对1985-2010年人均GDP指数拟合的RMSE为40.28，MAPE为1.30%。预测结果表明：“十二五”时期人均GDP指数仍将稳步增长，年均增速为9.35%，到2015年，人均GDP指数将超过6500。 3．采用GDP指数的建模方法，建立了河南省第一产业指数的半参数自回归模型。该模型对1984-2010年第一产业拟合的RMSE为19.45，MAPE为2.94%，且2011-2015年第一产业指数值分别为1111.1，1174.0，1221.5，1285.8，1347.4，其中对2011年的预测误差仅为0.32%，整个“十二五”时期第一产业指数的年均增速为4.61%。 4．基于多项式样条估计理论，对河南省第二产业指数建立了具有外生变量的半参数自回归模型，其中第二产业和第三产业的显著性滞后变量作为线性部分，第一产业作为非参数部分。该模型对1980-2010年第二产业指数拟合的RMSE为439.69，MAPE为3.5%。预测结果显示：“十二五”时期河南省第二产业指数由2011年的71683.4迅速增加到2015年的111376.9，年均增速为11.92%。 5．通过构建河南省第三产业指数的线性自回归模型，将非显著一阶滞后变量作为非参数函数变量，基于多项式样条方法建立了第三产业指数的半参数自回归模型。通过计算，所建立的半参数模型对1981-2010年第三产业指数拟合的RMSE为203.36，MAPE为2.49%，且“十二五”时期第三产业指数的年均增速为10.93%，除了2012年的环比增速为9.15%，其它年份环比增速都将超过10%。总之，“十二五”时期，河南省经济仍将呈现较为平稳的增长态势，人均GDP指数及三大产业指数也将稳步提高，其中第二产业在河南产业结构中仍居主导地位，第三产业也将得到较快的发展。
[Abstract]:Reasonable prediction of the economic development trend and situation in Henan Province during the 12th Five-Year Plan period provides quantitative basis for scientific decision-making and effective management of provincial party committee, provincial government and relevant departments, and has important practical significance for promoting the construction of Central Plains Economic Zone. The semi-parametric regression model contains both parametric and non-parametric components. It not only retains the advantages of the parametric model which is easy to interpret, but also has the flexibility of modeling. Based on the semi-parametric regression theory and method, this paper establishes a semi-parametric autoregressive model of Henan's gross domestic product (GDP) index, per capita GDP index and the index of first, second and third industries (100 in 1952). The details are as follows: 1. The linear autoregressive modeling process of GDP exponent in Henan Province is given in detail. The significant variable of the linear autoregressive model is the linear principal variable of the semi-parametric model, and the non-significant variable is the non-parametric functional variable. The semiparametric trend autoregressive model of GDP exponent is established by using polynomial spline estimation method. From 1985 to 2010, the mean square standard error (RMSE) of GDP index fitting is 61.45, and the average absolute percent error is 1.14. The forecast results show that the average annual growth rate of GDP index in Henan Province is 9.26 during the 12th Five-Year Plan period, which still presents a steady development trend. 2. Similar to the modeling method of GDP exponent, the semi-parametric autoregressive model of per capita GDP exponent is established, in which the variables of order 1, 3, 4 and 6 are linear principal variables, and the variables of order 5 lag are nonparametric functional variables. The RMSE of the model fitted to the per capita GDP index from 1985 to 2010 is 40.28 and 1.30. The forecast results show that the per capita GDP index will continue to grow steadily during the 12th Five-Year Plan period, with an average annual growth rate of 9.35. By 2015, the per capita GDP index will exceed 6500. 3. The semi-parametric autoregressive model of primary industry index in Henan Province is established by using GDP index modeling method. The RMSE of this model for the primary industry from 1984 to 2010 is 19.45% and 2.94 respectively, and the primary industry index values for 2011-2015 are 1111.1 / 1174.01221.5/ 1285.81347.4 respectively. The forecast error for 2011 is only 0.32, and the average annual growth rate of the first industry index is 4.61in the whole "12th Five-Year Plan" period. 4. Based on polynomial spline estimation theory, a semi-parametric autoregressive model with exogenous variables is established for the secondary industry index in Henan Province, in which the significant lag variables of the secondary industry and the tertiary industry are regarded as the linear part. Primary industry as a non-parametric part. The RMSE of the model fitting the secondary industry index from 1980 to 2010 is 439.69 and 3.5. The forecast results show that the secondary industry index of Henan Province increased rapidly from 71683.4 in 2011 to 111,1376.9 in 2015, with an average annual growth rate of 11.92 in the 12th Five-Year Plan period. 5. By constructing the linear autoregressive model of the tertiary industry index in Henan Province, the semi-parametric autoregressive model of the tertiary industry index is established based on the polynomial spline method, which takes the non-significant first-order lag variable as the non-parametric function variable. Through the calculation, the RMSE of the semi-parametric model fitting the tertiary industry index from 1981 to 2010 is 203.36 MAPE is 2.49, and the average annual growth rate of the tertiary industry index in the 12th Five-Year Plan period is 10.933. Except for the 9.15 percent growth rate in 2012, the other year's growth rate will exceed 10 percent. In a word, during the 12th Five-Year Plan period, Henan's economy will still present a relatively stable growth trend, and the per capita GDP index and the three major industrial indices will also steadily increase, and the secondary industry will still occupy a dominant position in Henan's industrial structure. The tertiary industry will also get faster development.
【学位授予单位】：河南科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F127;F224

【参考文献】