Study on the Export of BP Neural Network Model to China Based on Seasonal Adjustment

• Published in: Big Data Analytics for Cyber-Physical System in Smart City Vol. 1303; pp. 1672 - 1678
• Main Author: Qi, Ding
• Format: Book Chapter
• Language: English
• Published: Singapore Springer 2021; Springer Singapore
• Series: Advances in Intelligent Systems and Computing
• Subjects: AR-GARCH; ARIMA; BP neural network; Export forecast
• ISBN: 9789813345737; 981334573X
• ISSN: 2194-5357; 2194-5365
• DOI: 10.1007/978-981-33-4572-0_246

In this paper, the export volume, real exchange rate, China’s GDP, America’s IPI, and their seasonal variables are used as the determinants. Three methods, BP neural network, ARIMA, and AR-GARCH, are used to model and predict the export volume of China to the United States. Select the error-index, and compare the simulation results and prediction results of the three models with the real values. The results show that the three models are satisfactory. Although there are some differences in simulation and prediction ability, the ARIMA model has obvious advantages. This paper analyzes the causes of the above results and puts forward suggestions for improving China’s export based on the model.