对数据挖掘中关联规则算法的优化处理

Optimization algorithm of data mining association rules

在挖掘关联规则的执行过程中,早期循环生成最大项目集的过程是很重要的。文中提出基于哈希表的算法,对生成侯选项目集的过程进行了优化,尤其是时生成二维侯选项目集更是有效。由于在早期循环中,生成侯选项目集的势较小,使得能更有效地修剪数据库,从而减小了后期循环的计算代价,同时也减小了I/O请求。

In the mining association rules in the implementation of the early-cycle generating largest project the process is very important. In this paper, based on the hash table algorithm to generate the can- didate set the process has been optimized, especially when genera- ting two-dimensional candidate set is more effective. As in the eady cycle, generating the potential candidate set smaller, making more effective pruning database, thus reducing the computational cost of the late cycle, but also reduces the I / O requests.