Developing forecasting model of container flow among the major ports of Asia

• Main Authors: Huang, Jo-Wen, 黃若玟
• Other Authors: Tsai, Feng-Ming
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/32857992266210915941

碩士 === 國立臺灣海洋大學 === 航運管理學系 === 102 === In 2010, the amount of containers shipped in Asia accounts for more than 50% of the global containers, and more than 90% of the international trade in Taiwan is processed by ocean shipping. Therefore, the container flow information has become a critical issue for the port operators and liners to support their strategy plans and decision makings. According to the survey, the subscribed database can only provide container flows in terms of country to country and throughput at each port. In order to analyze port to port container flows among the major ports of Asia, this study applies artificial neural network and regression model to forecast container flows among major ports. First, use country to country container data to estimate port to port data that amount of import and export containers; second, use regression model to estimate the amount of import and export port to port container; third, use neural network model to predict the Port to Port container import and export volume. Estimate data and actual values predicted Hong Kong, Yokohama, Kobe and Nagoya port provided with the official website to do compared between the comparison results are similar, so estimation and prediction method of the present study is feasible.