Please use this identifier to cite or link to this item:
The ComparisStudy on the Prediction of Imported Contaon of the Seasonal Forecasting Models-A iner Volume for International Container Ports in Taiwan
Wen-Yi Peng;Ching-Wu Chu
|Contributors: ||NTOU:Department of Shipping and Transportation Management|
Classical Decomposition;Trigonometric Model;Regression Model with Seasonal Dummy Variables;and Grey Forecast
|Issue Date: ||2011-10-20T08:33:47Z
Abstract:To measure the profitability of a container port is based on the total volume handled in the harbor. In order to properly design, construct and manage a port, it is necessary to predict a port's expected volume in the port for the future. From the figures of monthly import volume of each container port in the recent four months, the trend of seasonal fluctuation was found. The data could be used to appropriately make a prediction on the volume of each port in the short-term. The purpose of this paper is to compare the accuracy of four forecasting models, i.e. Classical Decomposition, the Trigonometric Model, the Regression Model with Seasonal Dummy Variables, and the Grey Forecast, for the import volume of the international container ports in Taiwan. By using the method of verification on the actual data collected, we are able to prove which prediction model can provide the best accuracy. The research objective is set on the import volume of the three international ports of Keelung, Taichung and Kaohsiung. The testing data is derived from the monthly statistics of the import volume from Jan. 2001 to Dec. 2004. By comparing the findings based on the revaluation method, Mean Absolute Error (MAE), Mean Absolute Percent Error (MAPE) and Root Mean Squared Error (RMSE), the Classical Decomposition provided the most accurate predictions on the port of Keelung, where the Classical Decomposition and the Grey Forecast provided the most accurate predictions on Taichung, and the Grey Forecast provided the most accurate predictions on Kaohsiung ports. Different prediction mythologies each have their own merits and weaknesses, but, to be more practical, we have to find the most suitable method to fit our particular marine shipping industry needs of forecasting accuracy. Therefore, it is suggested to proceed studying the same outbound container volume, in order to find out whether the same result is obtainable as the study on imported cargo volume.
|Relation: ||15(2), pp.21-36|
|Appears in Collections:||[航運管理學系] 期刊論文|
Files in This Item:
All items in NTOUR are protected by copyright, with all rights reserved.