摘要
随着第三领域保险占有市场份额的日益增大,对该领域风险客户的研究分析显得更加重要。利用台湾某保险公司原始数据,运用聚类分析、决策树等方法建立高风险客户的判别模型,并对保险客户进行详细的分析,从而识别出高风险的客户,可以帮助保险公司避免不必要的风险损失,提高其盈利水平。
With more and more insurance of the third realm, research upon customers with risk is becoming important. This article will use some common data mining techniques, such as clustering and decision tree to model and analyze insurance customers, in order to distinguish ones with high risk. It will aim to give some suggestions to insurance institutions and avoid some unnecessary loss.
出处
《统计与信息论坛》
CSSCI
2009年第8期91-96,共6页
Journal of Statistics and Information
关键词
第三领域保险
数据挖掘
风险甑别
insurance status survey
data mining
risk recognition