A two‐stage forecasting approach for short‐term intermodal freight prediction

• DOI: 10.1111/itor.12337
• Publication Date: 2016-09-02T23:22:50Z
• Authors: José Antonio Mosc...; Ignacio Turias; Maria Jesús Jimén...; Juan Jesús Ruiz‐A...; María del Mar Cerbán
• Keywords: Port (circuit theory)

Abstract The forecasting of the freight transportation, especially short‐term case, is an important topic in daily supply chain management. Intermodal transportation subject to multiple complex calendar effects arising port environment. use prediction methods provides information that may be helpful as a decision‐making tool management and planning operations processes ports. This work addresses problem on basis by novel two‐stage scheme combination offer reliable predictions fresh weight Ro‐Ro (roll‐on/roll‐off) transport for 7 14 days ahead. study compares with weekly approach. applies database preprocessing Bayesian regularization neural networks (BRNN) Stage I. In II, ensemble framework best BRNN models used enhance I forecasting. results show assessed are promising predict time series transport.