摘要
在数据库和数据仓库中运用数据挖掘技术必须考虑挖掘系统的速度问题。当数据集大到相当程度时,挖掘工作只能在巨型机上进行;而由于系统的速度不够快,挖掘出来的知识将会是滞后的,它对决策支持不仅无效甚至是有害的。针对这一问题,提出了决策树算法的并行机制,并对并行性的性能进行探讨。
It is needed to think about the problem with speed-up when data mining techniques are applied to database or data warehouse. If datasets mined are large over a special level, they can only be handled on supercomputers. The information mined from history datasets is invalid and even negative if the speed of mining system is too slow. As far as this point is concerned, this paper presents parallel principle of decision tree algorithm and discusses the performance of its parallelism.
出处
《计算机工程》
CAS
CSCD
北大核心
2002年第8期77-78,共2页
Computer Engineering
基金
河北省自然科学基金资助项目()600225
