摘要
针对传统的个人信誉评估方法存在的缺陷,提出了一种基于K均值聚类和支持向量机结合的个人信誉评估方法。该方法先将测试数据集进行聚类,根据数据离聚类的数据分布来选取合适数据训练支持向量机,然后利用支持向量机进行分类。结果表明,同单一利用支持向量机分类进行比较,该方法减少了训练时间,同时具有较高的测试精度,比传统的个人信誉评估模型有更好的效果。
There are some problems exist in traditional individual credit assessment system. To solve those problems, a credit assessment model basesed on k-means method and support vector method is proposed. First the training samples are clustered using the K-means method. Then, the new samples defined according the feature of samples in cluster train the support vector machines, and to classify the test set by SVM. The result shows the approach improves training precision and test precision of the whole model compared with the traditional support vector classification method and improved the training speed.
出处
《信息技术》
2013年第2期42-44,47,共4页
Information Technology
基金
人工智能四川重点实验室开放课题资助项目(2009RY001)
关键词
个人信誉评估
K均值
支持向量机
聚类
速度
personal credit scoring
K-means
support vector machines
cluster
speed
