摘要
Apriori算法通过逐层搜索进行迭代,得到频繁项集,是获取关联规则最著名也是最原始的挖掘算法,但是该算法存在当数据量较大时效率较低、I/O操作的频繁出现等缺点。针对上述算法存在的两个缺点,本文提出了一种基于预判筛选多叉树的Apriori算法,通过将数据处理为数组的方式减少数据库扫描次数,构建频繁多叉树减少存储空间,并通过预判筛选的方式加快算法运行速度,从而达到对Apriori算法的优化。实验证明,该方法比原始的Apriori算法运行时间要短一些,效率得到提高。
Apriori algorithm gets frequent iterations by layer-by-layer searching, which is the most famous and original mining algorithm to obtain association rules. However, this algorithm has some shortcomings, such as low efficiency when the amount of data is large, frequent I/O operations and so on. In view of the two shortcomings of the above algorithms, this paper proposes an Apriori algorithm based on pre- screening multi-tree, which reduces the number of database scans by data processing as an array, builds frequent multi-tree to reduce the storage space, and speeds up the algorithm by pre-screening, so as to achieve the Apriori algorithm. Optimization of algorithm. Experiments show that this method is shorter than the original Apriori algorithm and the efficiency is improved.
出处
《信息技术与信息化》
2018年第10期69-72,共4页
Information Technology and Informatization
