论文部分内容阅读
在事务数据库中的周期性一般关联规则可以揭示类的不同层次之间的关系和呈现周期性变化。这些信息对于识别在关联中的趋势和预测非常有用。由于数据噪声对发现周期性一般关联规则的巨大影响 ,文中用噪声比来抑制数据噪声对发现周期性一般关联规则的影响。同时根据对周期性与一般高频集之间关系的分析 ,利用周期裁剪技术来节省挖掘时间 ,给出了 (Cyclic general-ized itemsets,CGI)算法。实验证明 ,该算法可高效地发现周期性一般关联规则。
Periodic general association rules in the transactional database can reveal the relationship between the different levels of the class and present periodic changes. This information is useful for identifying trends and forecasts in relationships. Due to the huge impact of data noise on the discovery of periodic general association rules, the noise ratio is used to suppress the effect of data noise on the discovery of periodic general association rules. At the same time, based on the analysis of the relationship between the periodicity and the general high frequency set, the periodic cutting technique is used to save the mining time, and the Cyclic general-ized itemsets (CGI) algorithm is given. Experimental results show that the algorithm can efficiently find periodic general association rules.