Title: Container throughput forecasting using non-additive forecast combination

Authors: Geng Wu; Yi-Chung Hu; Peng Jiang

Addresses: School of Economics and Management, Ningbo University of Technology, Ningbo 315211, Zhejiang Province, China; Department of Business Administration, Chung Yuan Christian University, Taoyuan 32023, Taiwan ' Department of Business Administration, Chung Yuan Christian University, Taoyuan 32023, Taiwan ' School of Business, Shandong University, Weihai 264209, China

Abstract: Containerisation is regarded as an important driving force of globalisation and international trade and has prompted the development of global ports. Accurately forecasting container throughput is crucial to maritime port planning and management. Previous studies have shown that combination forecasting not only has a higher predictive accuracy than that of a single model, but can also reduce the risk of failure during model selection. In this study, we propose a model of non-additive combination forecasting by using the fuzzy integral, to reflect the correlation between forecasting models. We used the quarterly container throughput of eight ports in China to test the proposed model. The empirical results were promising and indicated that the proposed model can outperform prevalent models of combination forecasting. This paper contributes to decision makers better understand expected demand for logistics of ports and adequately run for the long-term development planning and day-to-day operations.

Keywords: container throughput; fuzzy integral; port management; forecast combination.

DOI: 10.1504/IJSTL.2024.143136

International Journal of Shipping and Transport Logistics, 2024 Vol.19 No.2/3, pp.149 - 174

Received: 29 Oct 2022
Accepted: 30 Mar 2023
Published online: 04 Dec 2024 *

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