| 一种基于事务一项矩阵改进的FP—Growth算法 |
| |
| 引用本文： | 段西强,乔赛.一种基于事务一项矩阵改进的FP—Growth算法[J].泰安师专学报,2012(6):50-52. |
| |
| 作者姓名： | 段西强 乔赛 |
| |
| 作者单位： | 泰山学院信息科学技术学院,山东泰安271021 |
| |
| 摘 要： | FP—Growth算法是关联规则挖掘的一个经典算法。本文提出一种基于事务一项矩阵的挖掘算法MFP—Growth算法，利用向量矩阵表示事务数据库，整个挖掘过程只扫描一遍数据库，通过构造各个频繁项的条件矩阵，利用条件矩阵的向量运算来进行频繁模式的挖掘，减少了算法占用的内存空间，提高了挖掘效率．实验结果表明，MFP—Growth算法是有效可行的．
|
| 关 键 词： | 数据挖掘 FP—Growth 条件矩阵 支持度 |
An Improved FP - Growth Algorithm Based on Transaction - Term Matrix |
| |
| DUAN Xi -qiang,QIAO Sai.An Improved FP - Growth Algorithm Based on Transaction - Term Matrix[J].Journal of Taian Teachers College,2012(6):50-52. |
| |
| Authors: | DUAN Xi -qiang QIAO Sai |
| |
| Institution: | ( School of Information Science and Technology, Taishan University, Tai＇an, 271021, China) |
| |
| Abstract: | FP - Growth algorithm is one of the classical data mining algorithms. A new mining algorithm MFP - Growth based on transaction - term matrix is presented. Transaction database is replaced by vector ma- trix, only scanning once in mining procedure. MFP - Growth constructs conditional matrix of each frequent pattern, using operation of conditional matrix for mining frequent patterns, so the memory of the algorithm is reduced, mining efficiency is improved. Experiment result shows MFP -growth is effective and efficient. |
| |
| Keywords: | data mining FP - Growth conditional matrix support degree |
| 本文献已被 维普 等数据库收录！ |
|
