摘要
将基于数据挖掘的聚类分析和图形化属性分析方法引入海上交通流数据特性分析领域。阐述了数据挖掘的基本思想与方法,并着重介绍了使用开源数据挖掘工具WEKA对某海上航道的交通流数据进行的数据挖掘实验。实验综合使用聚类分析和图形化属性分析两种方法得到该航道的特性信息,其结果证明利用数据挖掘方法可得到有价值的交通特性信息,并能为海上交通的有效管理提供有力的决策依据。
The two datamining-based machine learning methods,clustering and graphical property analysis,are introduced to analyze characteristics of vessel traffic flow data.A new way is tried to implement vessel traffic data analysis making use of data mining technique.A similarity-based algorithm,K-Means,is selected in the clustering process for its simplicity and efficiency.The popular data mining tool WEKA is chosen to carry out experiments,the conclusion that clustering is a good way to generalize multi-factor related regulations is obtained according to the data mining results.
出处
《中国航海》
CSCD
北大核心
2009年第1期60-63,90,共5页
Navigation of China
关键词
水路运输
海上交通流
数据挖掘
聚类分析
图形化属性分析
算法
waterway transportation
marine traffic flow
data mining
clustering analysis
graphical property analysis
algorithm