连续属性的关联规则挖掘

摘要

In this paper, a new coding mode of genetic algorithm with three-segment chromosomes (individuals) is formulated for solving discovery of association rules from data with continuous attributes. This algorithm integrates discretization, reduction and mining association rules. In addition, niche technology is introduced into genetic algorithm to avoid premature phenomenon. The experiment shows that the algorithm is correct and effective.%对具有连续属性的数据进行关联规则挖掘,构造了一种新的编码方式,采用三段式染色体(个体),将连续属性离散化、属性约简和规则提取集成在一起,并将小生境引入到遗传算法中,以避免早熟现象.实验表明,该算法是有效的.