基于关联模式挖掘的决策规则提取方法  

Approach for decision rules generation based on association patterns mining

作　　者：贾桂霞[1] 张永[1] 

机构地区：[1]兰州理工大学计算机与通信学院

出　　处：《计算机工程与设计》2006年第12期2175-2177,2186,共4页Computer Engineering and Design

基　　金：甘肃省自然科学基金项目(3ZS042-B25-014)

摘　　要：在数据挖掘领域,关联规则的挖掘和基于粗糙集理论抽取决策规则是两种截然不同的方法,但在统计意义下两种方法产生的规则基本相同。结合关联规则挖掘方法和粗糙集方法的优点,基于Apriori算法提出一种优化算法,获取具有一定支持度和可信度阈值且不产生冗余的决策规则,以提高粗糙集属性值约简算法的性能。In data mining community, the methods of association rules mining and decision rules generation from the rough set model are strongly different. However, under statistical significance the both methods are basically identical with respect to derivation rules. An optimized method is presented to yield no redundant rules with certain support and confidence thresholds in which the advantages of association rule mining method-Apriori algorithm and rough set are unified. The method is expected to improve the performance of attribute value reduct based on rough sets.

关 键 词：数据挖掘 关联规则 粗糙集 APRIORI算法 决策表 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]