摘要
提出一种自适应的频繁模式挖掘算法:AD-Mine算法.该算法采用超结构,根据计算机可用内存自动确定一次性产生超结构的大小,能够自动适应各类不同特性的数据,进行高效率的频繁模式挖掘工作.同时提出了一种能够有效地减少扫描记录数的新颖的数据库划分方法.
An adaptive algorithm, AD-Mine, is put forward to mining frequent patterns. The algorithm can fit different conditions and achieve good performance. A alterable hyper-structure is used in AD-Mine. The algorithm can automatically adjust the size of the hyper-structure according to the memory available for accommodating different cases of data and mining frequent patterns effectively. A method of partitioning database is brought forward to reduce the counts of scanning records.
出处
《控制与决策》
EI
CSCD
北大核心
2004年第8期867-871,880,共6页
Control and Decision
基金
江苏省自然科学基金资助项目(BK2002091)
江苏省高校自然科学研究计划项目(03KJD110089).
