摘要
针对大多数关联规则数据挖掘算法难以适应支持度或数据集的变化问题,提出一种基于图像映射的关联规则数据挖掘算法Pix-DM。该算法利用图像在操作系统中的显示及存储特点,结合数据挖掘理论,通过映射有效地将数据挖掘过程在线性空间中实现,提高了算法对支持度或数据集变化的适应能力。实验证明,Pix—DM算法是有效且可行的。
Most data mining algorithms of association rules can not adapt well to the changes of support and datasets. This paper proposes a novel algorithm based on image mapping, namely Pix-DM. By using the points of images displaying and storage on operating system, and combining the method of data mining, it carries out the process of data mining in a linear space and improves the flexibility of the support and datasets changes. The experiment on real datasets proves that Pix-DM is effective and feasible.
出处
《计算机工程》
CAS
CSCD
北大核心
2008年第21期71-72,75,共3页
Computer Engineering
基金
国家"863"计划基金资助项目(2006AA04Z212)
关键词
数据挖掘
关联规则
图像
映射
data mining
association rule
image
mapping
