摘要
针对航运企业客户数据样本量少,信息量贫乏、不确定性等特点,提出了一种基于双向互动的粗糙-支持向量机的动态建模方法。利用该方法对具有复杂动态特性和不确定性的航运企业客户信用评估进行建模,并进行了实例研究。结果表明,基于粗糙-支持向量机的动态模型具有较快的收敛速度、较高的建模精度以及较好的泛化能力。
The integration of rough set theory and support vector machine is a new approach for complex nonlinear system modeling and provides a powerful way to process information for a complex nonlinear system with uncertain and incomplete data.With complementary advantages of rough set theory and SVM,the integration of the two is studied and a dynamic modeling method based on RSSVM is presented.The method is applied to construct a model of credit assessment of shipping enterprise customer having complex dynamic characteristics and uncertainties.It is proved by experiment that the method is superior to BP and MDA.
出处
《中国航海》
CSCD
北大核心
2011年第3期103-107,共5页
Navigation of China
基金
浙江省教育厅科研项目(Y200804313)
浙江省哲学社会规划项目(10HZCS05YB)
关键词
航运企业
支持向量机
粗糙集
信用评估
shipping enterprise
support vector machine
rough set
credit assessment
