一个新的不需要候选集的挖掘关联规则算法_Relim算法的研究_刘喜苹.pdfVIP

25 2 计 算 技 术 与 自 动 化 Vol125, No12 2 0 0 6 6 Computing T echnology and Automation Jun. 2 0 0 6 : 1003- 6199 2006 02- 0081- 04 Relim 1 2 2 刘喜苹 , 刘彩苹 , 谭义红 1. , 410082; 2. , 410082 : p- grow th 算法是当前挖掘频繁项目集算法中速度最快, 应用最广, 并且不需 候选集的一 种挖掘关联规则的算法但是, p- grow th 算法也存在着算法结构复杂和空间利用率低等缺点Relim 算 法是在 p- growth 算法的基础上提出的 一种新的不需 候选集的挖掘关联规则算法它具有算法结构简 单, 空间利用率高, 易于实现等显著优点本文在详细阐述Relim 算法后, 对 p- growth 算法和Relim 算法 的性能进行了分析和比较结果表明, Relim 算法尽管结构简单, 但其运行速度与 p- growth 算法相比并 不慢, 而且当对最小支持度高或者频繁规则比较少的数据集进行挖掘时, Relim 算法的运行速度往往比 p- growth 算法 快 :数据挖掘; 关联规则; 候选集; p- growth; Relim :TP301.6 :A Study on a New Mining Association Rule Without Candidate Generation% Relim Algorithm 1 2 2 LIU Xi-ping , LIU Ca-i ping , TAN Y-i hong 1.Software School of Hunan University, Changsha 410082, China; 2. Computer and Communication inst itute of Hunan University, Changsha 410082, China Abstract: p- growth algorithm is one of the currently fastest and most popular algorithms for mining association rule w ith- out candidate generation. However, it has disadvantages such as complicated data structure and lower space ut ilization rate. Recur- sive elimination Relim algorithm, which is proposed by Christian Borgelt in 2004 based on the p- grow th algorithm, is a new algorithm in mining association rule without candidate generation. The dramatic advantages of Relim algorithm are its simplicity of structure and high space utilization rate. In this paper, the Relim algorithm is described firstly, and t hen is compared w ith the p - growth algorit hm. T he comparison results show that, even though the Relim algorithm has simple structure, its speed is not slower than that of t he p- growt h algorithm, even faster when the minimal support is set high or there are few er frequent item sets. Key words:data mining; association rule; candidate generation; p- grow th; Relim : 1 , 2 1 [ 1] , Agrawal Apriori ; Park [ 2] DHP ; Savasere PART- [ 3] , T ITION , [456] : 2005- 10- 28 : 1978 ,