Article: Freight throughput forecast of China airport based on random forest algorithm Journal: International Journal of Shipping and Transport Logistics (IJSTL) 2025 Vol.20 No.3 pp.291 - 311 Abstract: In order to improve prediction accuracy and assist the air transport decision-making department in formulating a more reasonable transportation plan, this paper proposes the random forest algorithm (RFA) to handle high-dimensional variables and construct an airport freight throughput prediction model. Given that the airport's freight throughput is influenced by a complex environment, 18 influencing factors were assessed using Pearson correlation analysis to measure the strength of the correlation between these factors and airport freight throughput. The prediction model's critical parameters were then determined using grid search and cross-validation, including the number of decision trees, depth of decision trees, and the number of feature variables. The model's important variable scores were ranked, and optimal parameters were selected to establish the RFA regression prediction model. Finally, using Wuhan Airport as an example, the established RFA regression prediction model was used to verify the prediction results on the test set. The prediction effect was compared with the results of the autoregressive integrated moving average (ARIMA) model, multiple regression analysis, and the back-propagation (BP) neural network model. The results indicate that the RFA-based regression prediction model has higher prediction accuracy. Inderscience Publishers - linking academia, business and industry through research

Title: Freight throughput forecast of China airport based on random forest algorithm

Authors: Hang He; Jie Wang; Jinghui Zhang

Addresses: School of Economics and Management, Civil Aviation Flight University of China, No. 46 Nanchang Road, Guanghan, Sichuan, China ' School of Airport, Civil Aviation Flight University of China, No. 46 Nanchang Road, Guanghan, Sichuan, China ' Anzina Pty Ltd., Sydney, NSW, 2118, Australia

Abstract: In order to improve prediction accuracy and assist the air transport decision-making department in formulating a more reasonable transportation plan, this paper proposes the random forest algorithm (RFA) to handle high-dimensional variables and construct an airport freight throughput prediction model. Given that the airport's freight throughput is influenced by a complex environment, 18 influencing factors were assessed using Pearson correlation analysis to measure the strength of the correlation between these factors and airport freight throughput. The prediction model's critical parameters were then determined using grid search and cross-validation, including the number of decision trees, depth of decision trees, and the number of feature variables. The model's important variable scores were ranked, and optimal parameters were selected to establish the RFA regression prediction model. Finally, using Wuhan Airport as an example, the established RFA regression prediction model was used to verify the prediction results on the test set. The prediction effect was compared with the results of the autoregressive integrated moving average (ARIMA) model, multiple regression analysis, and the back-propagation (BP) neural network model. The results indicate that the RFA-based regression prediction model has higher prediction accuracy.

Keywords: airport freight throughput; random forest algorithm; RFA; Pearson correlation coefficient analysis; prediction; China.

DOI: 10.1504/IJSTL.2025.147397

International Journal of Shipping and Transport Logistics, 2025 Vol.20 No.3, pp.291 - 311

Received: 03 Jul 2023
Accepted: 23 Jan 2024
Published online: 15 Jul 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Freight throughput forecast of China airport based on random forest algorithm

Keep up-to-date