基于多最小支持度的挖掘算法研究  

Mining Algorithm Based on Multiple Minimum Supports

作　　者：常浩[1] 

机构地区：[1]太原大学计算机工程系,山西太原030032

出　　处：《太原大学学报》2013年第2期127-130,共4页Journal of Taiyuan University

摘　　要：数据挖掘是从事务数据库中抽取有用的知识和感兴趣的模式,而从事务数据库中发现关联规则是最常见的挖掘技术之一。提出一个遗传模糊关联规则挖掘框架和综合聚类、模糊和遗传概念的多最小支持度的遗传模糊关联规则挖掘算法。该算法从定量事务数据库中抽取合理的多最小支持度值、隶属函数和模糊关联规则,首先使用k-means聚类算法采集相似项目,然后初始化一个种群设定相同的支持度值,每一个染色体通过需求满足的标准和隶属函数的适应性来评估是否满足其适应度。Data mining is the process of extracting useful knowledge or interesting patterns from the transaction database. Find- ing association rules in transaction databases is one of the most common mining techniques in data mining. This thesis proposed a genetic fuzzy association rule mining framework and an integrated clustering, fuzzy and genetic concepts of multiple minimum supports genetic fuzzy association rule mining algorithm. The mining algorithm extracts reasonable minimum support value, member functions and fuzzy association rules from quantitative transaction database. Firstly, it uses the k-means clustering algo- rithm acquisition to collect similar projects, and then it initializes a support with similar population setting characteristics. Each chromosome assesses whether it meets the fitness needs by the standards and the adaptability of the member functions.

关 键 词：数据挖掘 遗传模糊算法 多最小支持度 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]