基于SARIMA模型和Holt-Winters平滑法的山西省进出口总额预测
Prediction of the Total Import and Export Volume of Shanxi Province Based on SARIMA Model and Holt-Winters Smoothing Method
摘要: 从国家宏观层面看,进出口总额体现了一国参与对外贸易的程度,是国民经济的关键性指标之一。而在我国的各个省份内,由于发展力度的差异,不同省份对对外贸易的依赖程度也不同。作为经济欠发达地区的山西,其对外贸易水平也在缓步增长。新冠疫情的到来,对进出口贸易也产生了一定的冲击。因此,分析山西省的进出口总额变化情况以及未来发展趋势尤为重要。本文以山西省2000~2021年进出口总额的264期月度数据为研究对象,通过建立、统计检验和拟合乘法SARIMA模型对其进行时间序列分析,并分析比较该模型和Holt-Winters指数模型的预测效果。研究结果表明,山西省月度进出口总额在时间上具有趋势性和季节性,并且模型拟合效果好。通过预测,为进一步制定贸易政策和优化贸易结构提供了理论依据。
Abstract: At the national macro level, the total import and export value reflects a country’s degree of partici-pation in foreign trade and is one of the key indicators of the national economy. In each province of our country, due to the difference in development intensity, the degree of dependence of different provinces on foreign trade is also different. As an economically underdeveloped area, Shanxi’s for-eign trade level is also growing slowly. The arrival of the new crown epidemic has also had a certain impact on import and export trade. Therefore, it is particularly important to analyze the changes in the total import and export volume of Shanxi Province and the future development trend. This pa-per takes the monthly import and export data of Shanxi Province from 2000 to 2021 as the re-search object, conducts time series analysis by establishing, statistical testing and fitting the multi-plicative SARIMA model, and analyzes and compares the prediction effect of the model and the Holt-Winters index model. The research results show that the monthly import and export volume of Shanxi Province has a trend and seasonality in time, and the model fitting effect is good. The fore-cast provides a theoretical basis for further formulating trade policy and optimizing trade structure.
文章引用:王瑞敏. 基于SARIMA模型和Holt-Winters平滑法的山西省进出口总额预测[J]. 应用数学进展, 2022, 11(9): 6454-6463. https://doi.org/10.12677/AAM.2022.119683

参考文献

[1] 王红, 童恒庆, 魏平. 进出口贸易额预测的逐步回归建模研究[J]. 统计与决策, 2006(2): 80-82.
[2] 李菊梅, 樊亮, 崔俊峰. 用时间序列分析方法预测我国的进出口总额[J]. 重庆工学院学报(自然科学版), 2008, 22(12): 136-138.
[3] 程兰芳, 陆敏. 中国服务贸易进出口规模的ARIMA模型和预测[J]. 黑龙江对外经贸, 2010(12): 25-27.
[4] 张玥, 杜宝军. 基于机器学习对我国进出口总额的研究[J]. 质量与市场, 2021(12): 150-152.
[5] 谭亮, 万丽娟. 中国对外直接投资与进出口贸易关系的实证分析[J]. 重庆大学学报(社会科学版), 2010, 16(1): 59-64.
[6] 胡登. 陕西进出口总额影响因素研究——基于多元线性回归模型[J]. 当代经济, 2018(14): 88-89.
[7] 俞力宁. 宏观经济因素对进出口贸易总额影响的研究——基于江苏省的实证分析[J]. 现代商业, 2022(9): 36-40. [Google Scholar] [CrossRef
[8] 李岚. 人民币汇率变动对我国进出口贸易影响的实证分析[J]. 现代营销(下旬刊), 2022(5): 4-7. [Google Scholar] [CrossRef
[9] 赵喜仓, 周作杰. 基于SARIMA模型的我国季度GDP时间序列分析与预测[J]. 统计与决策, 2010(22): 18-20. [Google Scholar] [CrossRef
[10] 戴玉泉. 基于ARIMA模型对江西省CPI的时间序列分析与预测[J]. 科学技术创新, 2019(35): 18-20.
[11] 强添纲, 刘涛, 裴玉龙. 地铁进站客流量SARIMA与GA-BP神经网络组合预测模型[J]. 铁道运输与经济, 2021, 43(12): 134-142. [Google Scholar] [CrossRef