基于ARIMA模型和指数平滑模型对我国货物进出口总额的预测研究

Prediction of China’s Total Import and Export of Goods Based on ARIMA Model and Exponential Smoothing Model

摘要: 在后疫情时代，为了研究我国货物进出口情况，本文采用了ARIMA模型和Holt-Winters三参数指数平滑模型对我国1996年1月至2022年10月的货物进出口总额的月度数据进行建模比较分析，并对未来三年的进出口总额进行预测。根据预测结果显示，两个模型都能较好的拟合数据，后者模型的拟合程度略优于前者。预测结果显示未来三年我国的货物进出口总额是逐渐增长的。

Abstract: In the post epidemic era, in order to study the import and export situation of China’s goods, this paper uses ARIMA model and Holt Winters three parameter exponential smoothing model to model and compare the monthly data of China’s total import and export of goods from January 1996 to October 2022, and forecasts the total import and export in the next three years. According to the prediction results, both models can better fit the data, and the fitting degree of the latter model is slightly better than the former. The forecast results show that China’s total import and export of goods will gradually increase in the next three years.