Models for Prediction of Global Trades by Port Container and Airport Cargo Throughputs

• Main Authors: Yu-Chin Yu, 尤郁晴
• Other Authors: Hua-An Lu
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/08838943128664761342

碩士 === 國立臺灣海洋大學 === 航運管理學系 === 99 === Trade development is the lifeblood of the country’s economic development and a composed indicator to measure a country’s GDP. In addition, trade development is one of the main reasons to promote the economic growth of country. Therefore, it is indeed necessary to estimate and forecast the trade data as well as provide prompt information, and then grasp the trend of global trade and economic development. Among numerous modes of transport, ocean shipping and air transport are the most two important modes for global trade. The reasons are that port container and airport cargo throughtputs are not difficult to control, and there is a time lag of International Trade Statistics published by WTO. This paper forecast the growth of global trade by grey forecasting model and stepwise regression. Using the grey relational analysis to eliminate and select four ports and three airports with development trends closer to the value of global trade. As a result, the selected ports are Bremen/Bremerhaven, Ningbo, Singapore and Tianjin Port; three airports are Paris, Louisville and Frankfurt Airport. In grey forecasting, we fist use grey GM(1,1) forecasting theory to establish the model. Second, establish GM(1,5) and GM(1,4) by four ports and three airports and then analyze the accuracy of models. The purpose is to verify which grey forecasting results of model is the most appropriate way in forecasting global trade. The results indicated that the performance of GM(1,5) is the best suitable model on global trade forecast, followed by airport CDG-LAX-LHR model of step regression.