摘要: 提出一种通过分组达到并行化的改进Apriori算法,研究了分组的构造以及分组后因连接步和剪枝步的简化而达到的效率的提高。对于算法中涉及的原理给予了完整的论证并结合实例给出了改进算法的实现步骤。结果表明,改进算法通过连接步和修剪步的简化大大地提高了规则发现的效率,并可以很方便地达到并行化。
关键词:
关联规则,
分组,
并行化,
频繁集,
候选集
Abstract: The paper puts forward a improved Apriori arithmetic which achieves parallel processing via grouping,studies the construction of the grouping and the improvement of the efficiency for the predigestion of the join and pruning after grouped. For the principle relate to in the arithmetic , it gives the full prove and realization step of the improved arithmetic by linking the instance. The results show that the improved arithmetic improves the efficiency of discovering rules by the predigestion of the join and pruning, and can achieve paralleling conveniently.
中图分类号:
张伟,郑涛,李辉. 一种并行化的分组关联规则算法[J]. , 2004, 30(22): 84-.
ZHANG Wei,ZHENG Tao,LI Hui. Parallel Mining Association Rules Based on Grouping[J]. , 2004, 30(22): 84-.