东北大学学报（自然科学版）

东北大学学报(自然科学版) ›› 2006, Vol. 27 ›› Issue (2): 153-156.DOI: -

• 论著 • 上一篇    下一篇

一种基于聚合链的改进FP-Growth算法

焦明海;姜慧研;唐加福;   

  1. 东北大学计算中心;东北大学计算中心;东北大学信息科学与工程学院 辽宁沈阳110004;辽宁沈阳110004;辽宁沈阳110004
  • 收稿日期:2013-06-23 修回日期:2013-06-23 出版日期:2006-02-15 发布日期:2013-06-23
  • 通讯作者: Jiao, M.-H.
  • 作者简介:-
  • 基金资助:
    辽宁省自然科学基金资助项目(20042020)

Improved FP-growth algorithm based on aggregative chains

Jiao, Ming-Hai (1); Jiang, Hui-Yan (1); Tang, Jia-Fu (2)   

  1. (1) Computer Center, Northeastern University, Shenyang 110004, China; (2) School of Information Science and Engineering, Northeastern University, Shenyang 110004, China
  • Received:2013-06-23 Revised:2013-06-23 Online:2006-02-15 Published:2013-06-23
  • Contact: Jiao, M.-H.
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

637

摘要： 提出了一种基于聚合链挖掘频繁模式的改进FP-growth算法.该算法引入聚合链的单链表结构,改进了FP树结构.改进后的FP树是单向的,每个结点只保留指向父结点的指针,节省了树空间;相同项的不同节点的路径信息压缩进聚合链中,避免了生成节点链和条件模式库.用Agrawa方法生成实验数据进行分析,实验结果验证了该算法在时间上的优势.

关键词: 数据挖掘, 频繁模式, FP树, 聚合链, FP-growth算法

Abstract: An improved frequent pattern FP growth algorithm based on aggregative chains is proposed. A kind of single linked lists named aggregative chain is introduced to the algorithm, thus improving the architecture of FP tree. The new FP tree is a one-way tree and only the pointers to point its children at each node are kept to save the space of tree in comparison with the former one. Route information of different nodes in the same term are compressed into aggregative chains so that the frequent patterns will be produced in aggregative chains without generating node links and conditional pattern bases. Agrawa tests data verified the advantage of time occupancy of the proposed algorithm.

中图分类号: