基于数据挖掘技术的信用卡审批模型研究  

Research on credit card approval models based on data mining technology

作　　者：花蓓[1] 

机构地区：[1]南华工商学院计算机系,广东广州510507

出　　处：《计算机工程与设计》2008年第11期2989-2990,F0003,共3页Computer Engineering and Design

摘　　要：针对分类预测建模数据的非对称性,提出一种基于神经网络和决策树技术结合的非对称性数据集合预测分类建模方法,建立了信用卡审批模型。结果表明:增加预测类标识决策属性后,在用不同比例的建模数据集建立的所有模型中,比例为33.33%:66.67%的数据集建立的神经网络模型最好,模型的准确率达到88.49%。Because the unbalance of the data sets of building model would affect the credit card classified prediction, it is raised that the predictive classified building model method of unbalance data sets is based on the combination of decision tree and neural net and the model of credit card approval are built. The result shows that after added predict field among all models which are built using different data sets, the neural net model with the 33.33%： 66.67% data set has the best performance. The accuracy rate of the model is 88.49%.

关 键 词：数据挖掘 信用卡 神经网络 决策树 模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]