摘要
基于TEI@I理论框架,本文提出了适用于港口物流货运复杂系统的TEI@I的综合集成预测模型,并基于青岛港集装箱吞吐量数据,预测和分析了TEI@I的港口物流货运量的集成预测理论框架的各部分.预测结果表明,基于TEI@I集成预测模型的效果远远优于单独模型的预测结果,方向变化统计量(Dstat)的评价结果从66%-77%提高到了100%,对主要决策者方向性的判断更有实际意义.其次,TEI@I方法论中将数据"先分解后集成"的思想,引入了对数据的非线性部分的分析和预测,该方法不仅提供了分析外部冲击对具体数据序列的影响程度,影响周期的分析思路,而且将分析后的序列集成,对不同模型的预测精度有了很大的提高.
This paper presents an integrated forecasting model based on the TEI@ I methodology for port logistics prediction. For illustration, Qingdao port container throughout series is used as a case study. Empirical results reveal that TEI@ I integrated model can significantly improve the prediction performance over single models presented in this study, especially in terms of direction prediction measurement (Dstat). It implies that the integrated forecasting model based on the TEI@ I methodology can be used as a feasible solution to port logistics volume prediction and analysis. Furthermore, it also can be applied as an effective and feasible solution to analyze the effects of irregular and infrequent events on the time series of forecasting object.
出处
《交通运输系统工程与信息》
EI
CSCD
北大核心
2012年第1期173-179,共7页
Journal of Transportation Systems Engineering and Information Technology
基金
国家自然科学基金重点项目(NSFC Grant No.70810003)