关联规则算法的实现与改进

关联规则作为一种数据挖掘的工具,它能够发现数据项集之间有趣的关联。在关联规则的算法中,Apriori算法是其中的关键算法之一。面对大量复杂的数据集,怎样选择数据结构,怎样优化处理过程,对于此算法的性能将会十分重要。该文首先介绍了关联规则的原理和Apriori算法的实现,然后提出了对该算法的若干改进,例如:采用树型结构存取频繁项集,使用三种缓存优化的方法等。这些优化都能够在整体上提高算法的效率。对于大数据项,试验显示,这些改进能够正确、有效、快速地实现Apriori算法。

Realization and Optimization of Association Rule Mining Algorithm

Association rule mining searches for interesting relationships among items.Association rule is now used as one of efficient data mining tools.In all association rule algorithms ,the apriori algorithm is one of the keys.To huge scale and complex of data sets,how to choose the data structure,how to optimize the process,are all very important for the performance of the algorithm.This paper first introduces the principle of association rule and its realization by apri-ori algorithm.Then some improvements of apriori algorithm are proposed,such as creating a tree structure to store all frequent itemsets and using three kinds of cache optimization methods.These optimization can improve the algorithm ef-ficiency at the whole.The experiments show these improvements proposed in this paper can realize the Apriori algorithm accurately,efficiently and quickly in large data sets.