改进的基于FP-tree的频繁项集挖掘算法  

Improved algorithm for frequent itemset mining based on FP-tree

作　　者：郭伟[1] 叶德谦[1] 

机构地区：[1]燕山大学信息工程学院,河北秦皇岛066004

出　　处：《计算机工程与应用》2007年第19期174-176,共3页Computer Engineering and Applications

基　　金：河北省自然科学基金(the Natural Science Foundation of Hebei Province of China under Grant No.F2004000252);教育部留学回国人员科研基金(No.2001498)

摘　　要：对于频繁项集挖掘,采用一种FP-数组技术来减少FP-tree的遍历时间,减少数据集的扫描次数,在此基础上提出了一种基于FP-tree进行频繁项集挖掘的FP-growth+算法,提高了算法的效率。最后的实验证明了该算法的有效性。Using the FP-array technology to reduce the time of FP-tree traversal,the performance of algorithm FP-growth is improved as for frequent itemsets mining because of the decreasing time.An algorithm FP-growth＋ based on FP-tree for mining frequent itemsets is proposed.It can improve the efficiency of the algorithm.Lastly the algorithm is realized on experiment and is proved to be valid.

关 键 词：数据挖掘 关联规则 频繁项集 FP—tree 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]