基于改进Apriori算法的决策推导过程

Mining decision rules based on the improved Apriori algorithm

计算机的广泛应用,使得人类积累了海量的数据。利用数据挖掘技术,可以从大型数据库或数据仓库中提取出隐含的,未知的及有潜在应用价值的信息或模式。关联规则是数据挖掘中最活跃的分支之一,侧重于挖掘数据库中各数据项间隐藏的深层次的关联关系,分析出潜在的行为模式。Apriori算法是挖掘关联规则的最经典的算法。文中介绍了经典Apriori算法的基本方法,并从数据项建立,频繁项集连接,规则生成3个方面对Apriori进行改动,挖掘出了可用于决策的规则。

The wide range application of the computer accumulated vast amounts of data. Data mining can extracted the implied, unknown, and the potential value of information or patterns from large database or data warehouse. Association rules is one of the most active branch of da- ta mining, focusing on mining the deep-level relationship among data items in the database, and analyzing the potential behavior patterns. Apriori algorithm is the most classical algorithm for mining association rules. This paper introduced the basic method of Apriori algorithm, and changed it in three aspects： data item establishment, frequent item sets connection and the rule generation. Using the changed algorithm, mined rules can be used for decision-making.