摘要
以提高信用等级评价的质量为目的,介绍了数据挖掘技术的基本过程.以企业贷款的信用分类为研究背景,具体研究了业务理解、数据理解、数据准备、建模、评估和发布的实现环节.在建模过程中,采用决策树为分析模型,对经典的C4.5算法进行了改进.将改进算法运用在企业贷款的信用分类中,并将其效果与经典的C4.5算法的结果进行比较,结果表明该算法对于企业信用分类这样的复杂系统,在准确度与决策树结构上具有一定程度上的改善,能够提高信用等级评价质量.
Aiming at the improvement of credit grade evaluating quality, the basic processes of data mining technique are introduced. Taking credit classifying of enterprise loans as research background, implement links such as business understanding, data comprehension, data preparation, modeling, evaluation and release were studied. During modeling, the decision tree was adopted as analyzing model, and the conventional CA.5 algorithm was improved. The new algorithm was applied in credit classifying of enterprise loans, and its effect was compared with that of conventional C4.5 algorithm. The result shows that the improved algorithm can increase precision and improve structure of decision tree in complex credit classifying, which is capable of improving the quality of credit grade evaluation.
出处
《沈阳工业大学学报》
EI
CAS
2007年第6期685-691,共7页
Journal of Shenyang University of Technology
基金
国家自然科学基金资助项目(70671016)
关键词
数据挖掘
信用等级
决策树
C4.5算法
分类
data mining
credit grade
decision tree
C4.5 algorithm
classification
