基于改进多层次模糊关联规则的定量数据挖掘算法

针对单一层次结构实现规则提取具有规则提取准确性不高、算法运行时间长、难以满足用户使用需求的问题，提出一种基于改进多层次模糊关联规则的定量数据挖掘算法。采用高频项目集合，通过不断深化迭代的方法形成自顶向下的挖掘过程，整合模糊集合理论、数据挖掘算法以及多层次分类技术，从事务数据集中寻找模糊关联规则，挖掘出储存在多层次结构事务数据库中定量值信息的隐含知识，实现用户的定制化信息挖掘需求。实验结果表明，提出的数据挖掘算法在挖掘精度和运算时间方面相较于其他算法具有突出优势，可为多层次关联规则提取方法的实际应用带来新的发展空间。

Quantitative data mining algorithm based on improved multi-level fuzzy association rules

The accuracy of the rule extraction is not high, the algorithm runs long, and it is difficult to meet the needs of the users in extracting rules from a single hierarchy. To solve the above problems, this paper proposed a quantitative data mining algorithm based on the improved multilevel fuzzy association rules. It adopted the high frequency project set, formed the mining process by continuous deepening of the iterative method. This method integrated fuzzyset theory, data mining algorithm and multi-level classification technology to find fuzzy association rules from the transaction data set, and excavated the hidden knowledge of quantitative value information in the multi-layer structured transaction database. It realized the user''s customized information mining needs. The experimental results show that the quantitative data mining algorithm based on the improved multilevel fuzzy association rules has a prominent advantage over other algorithms in mining precision and operation time. It can bring new development in the practical application of multilevel association rule extraction.