关联规则挖掘Apriori算法的研究与改进

关联规则挖掘是数据挖掘研究领域中的一个重要任务,旨在挖掘事务数据库中有趣的关联.Apriori算法是关联规则挖掘中的经典算法.然而Apriori算法存在着产生候选项目集效率低和频繁扫描数据等缺点.对Apriori算法的原理及效率进行分析,指出了一些不足,并且提出了改进的Apriori_LB算法.该算法基于新的数据结构,改进了产生候选项集的连接方法.在详细阐述了Apriori_LB算法后,对Apriori算法和Apriori_LB算法进行了分析和比较,实验结果表明改进的Apriori_LB算法优于Apriori算法,特别是对最小支持度较小或者项数较少的事务数据库进行挖掘时,效果更加显著.

RESEARCH AND IMPROVEMENT OF APRIORI ALGORITHM FOR MINING ASSOCIATION RULES

Mining association rule is an important task in data mining research field,its purpose is to mine interesting associations in transaction database.Apriori algorithm is a classical algorithm for mining association rule.However,it has some disadvantages such as inefficient in generating candidate itemsets and frequently scanning database,etc.Based on the study of principle and efficiency of the Apriori algorithm,in this paper it points out its defects and presents an improved Apriori_LB algorithm.The new algo...