摘要
如何找出保险业务数据中有关投保和理赔的规律是保险公司能否提高盈利的至关重要的问题。关联规则发现是数据挖掘技术的一种简单又很实用的方法。文章首先介绍了关联规则的定义以及关联规则的 4个属性 :可信度、支持度、期望可信度和作用度。然后讲述了如何用 SGI公司的数据挖掘工具 Mineset在保险业务数据中发现关联规则 ,从而得出一些对保险公司起指导作用的控制投资风险的规则。关联规则还可广泛用于银行、电信。
Insurance companies need to find the rules for applications and claims in insurance business data to make a profit. Mining for association rules is a simple and very useful data mining method. This paper introduces the definition of the association rule and it's four attributes: confidence, support, expected confidence and lift. Then some association rules are mined with SGI (Silicon Graphics Incorporation) Mineset, a data mining tool. Some risk control rules are given that play important roles in the insurance company business. The association rules are widely used in the field of banking, electronic communication, commerce, etc.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2001年第1期45-48,共4页
Journal of Tsinghua University(Science and Technology)
关键词
数据挖掘
关联规则
可信度
支持度
data mining
association rule
confidence
support
decision support
