摘要
鉴于个人贷款在银行业务中的比重不断上升,对个人贷款申请进行审批显得格外重要。由于初始规则库常常存在知识不完全或不一致等问题,所以需要规则库修正,即规则库知识求精。现以日本银行个人贷款审批规则库为例,采用神经网络技术,利用个人贷款审批的历史数据,对规则库进行知识求精。结果表明,求精后的规则数量减少了,而分类准确率却有很大的提高。
It is important to approve application to private credits in banks, as private loans are becoming a principal service in bank operations. The approving rule base for private credit should be leaned after some time. Taken Japanese credit screening rule base as an example, the refinement procedure for rule base refinement was illustrated by using artificial intelligence technology and history credit screening data. The results show that the numbers of leaned rules decrease while the accuracy of the classification increases.
出处
《管理学报》
2006年第5期529-532,共4页
Chinese Journal of Management
基金
国家自然科学基金资助项目(70271002)
关键词
神经网络
交叉熵
规则库求精
个人贷款审批
neural network
crossentropy
rules base refinement
private credit of approving
