基于改进FP-Tree的最大频繁项集高效挖掘算法  

Efficient Algorithm for Mining Maximum Frequent Item Sets Based on Improved FP-Tree

作　　者：纪怀猛[1] 

机构地区：[1]福州大学阳光学院,福州350015

出　　处：《计算机与数字工程》2014年第6期959-963,共5页Computer & Digital Engineering

摘　　要：基于FP-Tree的FP-Max算法在挖掘最大频繁集时需多次递归建立条件模式树耗费大量存储空间,这大大降低了算法的挖掘效率。提出了一种基于改进FP-Tree的最大频繁集快速挖掘算法—FP-EMax算法。该算法无需建立条件模式库大大减少了存储空间开销,采用预剪枝策略减少条件模式树的构造次数及子集检测次数,从而算法的挖掘效率大大提高。最后通过实验证明FP-EMax算法在支持度较小的情况下较之于FP-Max及同类算法具有更好的性能。The efficiency of the algorithm for mining maximum frequent set is greatly reduced ,when the algorithm named FP-Max based on FP-Tree is used in mining maximum frequent item sets .Because it needs recursively to establish conditional pattern tree ,and take a lot of storage space .A new algorithm named FP-EMax based on improved FP-Tree is put forward for efficiently mining maximum frequent sets .The efficiency of the new algorithm is greatly increased ,because not only it doesn＇t need to establish conditional pattern library ,which greatly reduces the storage space overhead ;but also it uses pruning to reduce the number of establishing conditional pattern tree and testing subset .Finally ,the experiments show that the algorithm has a better performance than the FP-Max algorithm and other similar algorithms in the case of a relatively small support .

关 键 词：递归 最大频繁集 FP-TREE 条件模式树 索引 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]