摘要
数据挖掘算法对于支持度改变及数据集更新的适应性一直都是一个难点。本文根据数据集逐步增加的增量式方法,映射事务模式于线性空间中进行挖掘,并借助了图像在操作系统中显示及存储的特点,提出了一种新的增量式数据挖掘算法IPM-DM。同时,对比分析了其中两种模式映射方法,并与同类算法比较,经过实验证明,算法IPM-DM是有效且可行的。
The data mining algorithms are difficult adapted to the different supports and the updating datasets. Based on the incremental approach of dataset, a novel incremental algorithm IPM-DM of data mining is proposed. It maps the transaction patterns in linear space, and it uses the points of images displaying and storage on operator system. At the meanwhile, two pattern mapping methods of IPM-DM are compared, and algorithm IPM-DM is also compared with congeneric algorithm. The experiment proved that IPM-DM is effective and feasible.
出处
《齐齐哈尔大学学报（自然科学版）》
2009年第4期30-34,共5页
Journal of Qiqihar University(Natural Science Edition)
关键词
数据挖掘
关联规则
增量式
模式
映射
data mining
association rule
incremental
pattern
mapping
