CAJ | 학술논문

Apriori算法是关联规则挖掘中的经典算法，一直是数据挖掘领域的研究热点。传统的Apriori算法由于产生过多的无用的候选项集以及需要多次扫描数据库导致在一定程度上限制了算法的效率。本文针对这一问题，提出一种新的RF－Apriori算法。该算法首先对数据进行二元处理；然后利用项集的反单调性减少候选项集的产生，从而提高算法效率。实验结果表明，RF －Apriori算法效率明显优于Apriori算法。
Apriori산법시관련규칙알굴중적경전산법，일직시수거알굴영역적연구열점。전통적Apriori산법유우산생과다적무용적후선항집이급수요다차소묘수거고도치재일정정도상한제료산법적효솔。본문침대저일문제，제출일충신적RF－Apriori산법。해산법수선대수거진행이원처리；연후이용항집적반단조성감소후선항집적산생，종이제고산법효솔。실험결과표명，RF －Apriori산법효솔명현우우Apriori산법。
Apriori algorithm is a classical association rule mining algorithm,which has been a hot topic in the field of data mining.Because of the traditional Apriori algorithm generating excessive useless candidate items and scanning the database repeatedly the algorithm is limited for its efficiency.Aiming at this problem,we propose a new RF -Apriori algorithm.Firstly,data is converted into bivariate;Then use the anti -monotonic nature of itemsets is used to reduce candidate sets ,thus improving efficiency of the algorithm.Experimental results show, RF-Apriori algorithm is more efficient than Apriori algorithm.