一种基于关联模式的完全决策规则的提取方法  

A method for complete decision rules extraction based on association patterns

作　　者：贾桂霞[1] 张永[1] 陈思睿[1] 

机构地区：[1]兰州理工大学计算机与通信学院,甘肃兰州730050

出　　处：《兰州理工大学学报》2006年第5期104-107,共4页Journal of Lanzhou University of Technology

基　　金：甘肃省自然科学基金(3ZS042-B25-014)

摘　　要：针对基于粗糙集理论提取完全的决策规则是NP难问题,提出一种获取统计意义下的完全的简化规则的算法.该算法结合关联规则挖掘方法和粗糙集方法的优点,从决策表中提取出具有一定支持度和可信度阈值的决策规则,具有实际的应用意义.通过实例验证了算法的有效性.It is a NP difficult problem to generate complete decision rules based on rough set theory. An algorithm was proposed to produce complete and simple rules in a sense of statistical significance, in which the advantages of both association rule mining method and rough set were unified so, that the decision rules with certain support and confidence thresholds were extracted from decision table, being of application meaning. The validity of this algorithm was verified by an illustration.

关 键 词：数据挖掘 关联规则 粗糙集 APFIORI算法 决策表 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]