摘要
针对传统信用评价方法多是静态评价的不足,本文提出了一种融合SOM与K-means算法的动态信用评价方法。文章首先对动态信用评价问题进行了介绍,并利用E-TOPSIS方法对单时点下的静态信息进行集结,以确定被评价对象的信用评价值;然后在融合SOM算法和K-means算法各自优势的基础上,提出了SOM-K算法的原理和步骤;最后以SOM-K算法对被评价对象进行聚类,并确定相应信用等级。文章最后进行了实例验证。验证结果表明,该方法能够较好地克服静态信息下由于信息突变造成评价结果失真的问题。
Considering that traditional credit evaluation methods mostly belong to the category of static credit eval -uation , a dynamic credit evaluation method which integrates SOM and K-means algorithm is proposed to deal with this problem.Firstly, the paper introduces the basic model of dynamic credit evaluation , and employs E-TOPSIS to cluster the static information at single time points .Next, combining advantages of both SOM algorithm and K-means algorithm, the paper puts forward the principle and steps of SOM-k-means algorithm;then the integrat-ed clustering algorithm is utilized to cluster the evaluation alternatives and determine the corresponding credit rating of them.Finally, to verify the effectiveness of the method proposed, an example is employed.The example shows that the problem of information mutation under static situation can be greatly solved by the method proposed .
出处
《运筹与管理》
CSSCI
CSCD
北大核心
2014年第6期186-192,共7页
Operations Research and Management Science
基金
国家自然科学基金资助项目(71361021
71001048)
江西省科技厅科技资助项目(GJJ14113)
江西省社会科学"十二五规划项目"(13GL38)
江西省"赣鄱英才555"工程项目
国家社科基金项目(11BGL063
12FJL002)