摘要
传统关联规则挖掘算法往往产生过多规则而难以被决策者所采用。针对该问题,文章从应用的角度提出了最简关联规则,其特点是后项只包括一种产品,同时追求规则前项产品项数的最小化,在此基础上给出了一种最简关联规则挖掘算法。利用该算法得到的最简关联规则集包括的规则数量大为减少且能得出与全部关联规则集相同的决策,避免了大量的冗余挖掘,提高了挖掘效率和应用效果。
Conventional mining algorithms often produce too many rules for decision makers to digest. Instead, the concept of minimal association rules is introduced from the aspect of application in this paper. Minimal rule set, which includes rules with single item as consequent and the minimal number of items as the antecedent, can be used to derive the same decisions as other association rules without information loss, while the number of minimal rules is much less than of all rules. A mining algorithm without redundant rules is proposed and the mining efficiency is improved.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第13期46-48,共3页
Computer Engineering
基金
中国博士后科学基金资助项目(20060400705)
河北省科学技术研究与发展基金资助项目(054572248)
关键词
数据挖掘
关联规则
最简关联规则
data mining
association rule
minimal association rule
