挖掘关联规则算法的优化处理

Algorithm Optimization of 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，尤其是对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。

To find all the large itemsets from candidate sets in eary iterations is usually the domaining factor foroverall data-mining performance. In the paper,we option the algorithm Apriori for the candidate set generation. It is ahash-based algorithm and is especially effective for the generation of candidate set for large 2-itemsets. Furthermore thegeneration of smaller candidate sets enables us to effectively trim the transaction database size at a much earier stageof the iterations,thereby reducing the computational cost for later iterations. The advantage of proposed algorithm alsoprovides us the opportunity of reducing the amount of disk I/O required.