半参数可加自回归模型在我国GDP预测中的应用  

Application of Semi-Parametric Additive Autoregression Model in China’s GDP Forecast

作　　者：谢成兴 王丰效 聂僮 XIE Cheng-xing;WANG Feng-xiao;NIE Tong(School of Mathematics and Statistics,Kashi University,Kashi 844000,Xinjiang,China;College of Biological Sciences,China Agricultural University,Beijing 100093,China)

机构地区：[1]喀什大学数学与统计学院,新疆喀什844000 [2]中国农业大学生物学院,北京100193

出　　处：《兰州文理学院学报（自然科学版）》2021年第5期30-36,共7页Journal of Lanzhou University of Arts and Science(Natural Sciences)

基　　金：国家社会科学基金项目(11XTJ001);喀什大学研究生科研创新项目(KS2020010)。

摘　　要：GDP时间序列数据具有线性和非线性特征.自回归模型能解释线性部分,而单纯地使用线性回归分析GDP时间序列数据不符合实际情况.将自回归模型中不具显著性的解释变量作为非参数部分引入自回归模型,充分考虑GDP时间序列的线性和非线性的混合特点,构建半参数可加自回归模型,该模型更符合现实意义.利用我国1978~2019年的GDP数据进行实证分析,结果表明该模型不仅反应了GDP时间序列的非线性特征,而且在解释能力、拟合效果和预测精准度上有一定优势.GDP time series data has linear and nonlinear characteristics.Autoregression model can explain the linear part,but simply using linear regression analysis of GDP time series data is not in line with the actual situation.The non-significant explanatory variables in the autoregressive model are introduced into the autoregressive model as a non-parametric part.Considering the linear and nonlinear mixing characteristics of GDP time series,a semi-parametric additive autoregressive model is constructed,which is more realistic.Using the GDP data of China from 1978 to 2019,the results show that the model not only reflects the nonlinear characteristics of GDP time series,but also has certain advantages in interpretation ability,fitting effect and prediction accuracy.

关 键 词：自回归模型 半参数可加自回归模型 GAM GDP预测 

分 类 号：F224.9[经济管理—国民经济]