基于三段式编码GA的连续属性关联规则挖掘方法

Association Rules Mining Method from Data with Continuous Attribues Based on Three-sectional Coding GA

如何有效地从具有连续属性的数据中挖掘关联规则,是目前数据挖掘领域的一个研究热点。基于遗传算法,提出了一个连续属性关联规则挖掘方法。在该方法中,首先采用三段式编码将连续属性离散化、属性约简和规则提取集成在一起,然后将小生境引入到遗传算法中,以避免早熟、提高挖掘效率。实验表明,该方法是有效的。

At present, it has become a hotspot for data mining that how to effectively mine association rules from data with continuous attributes. A relevant association rules mining method based on GA is proposed. The pro- posed method firstly integrates discretization on continuous attributes, attribute reduction and rule acquisition to- gether by means of three-sectional coding, and then employs niche to avoid precocious phenomena and improve mining efficiency. The experiment results show that the method is effective.