摘要
本文以汪寿阳(2004)提出的TEI@I方法论为指导,提出了香港集装箱吞吐量预测的一种研究框架。采用季节AR IMA和VAR等计量经济模型、径向基神经网络技术,以及不规则事件的量化方法,建立了一个香港集装箱吞吐量的综合集成预测模型。数据实证显示这种预测模型具有较好的预测精度和稳健性。此外,通过VAR模型的建立和G ranger因果检验,指出沿海港口的发展对香港有重要的影响,香港港口发展应该注意区域竞争和协作问题。
Based on the TEI@I methodology proposed by Wang (2004), a forecasting framework about Hong Kong's container throughput is proposed in this paper. Applying Seasonal-ARIMA model, VAR model, RBF neural network technology and quantitative method with irregular events, we build an integrated forecasting model for Hong Kong' s container throughput. Empirical analysis shows that it performs fairly well in both precision and stability. Furthermore, after building VAR model and performing the Granger test, we observe that the development of coastal ports has a significant influence on Hong to regional competition and collaboration. Kong, and thus we suggest that attention should be paid
出处
《运筹与管理》
CSCD
北大核心
2009年第4期82-89,共8页
Operations Research and Management Science
基金
国家自然科学创新研究群体基金资助项目(70221001)
关键词
数量经济学
基于TEI@I方法论预测
人工神经网络
区域竞争
集装箱吞吐量
quantitative economics
forecasting with TEI@ competition
container throughput I methodology
artificial neural network
regional