区间值关系数据库上模糊关联规则的预测方法

Prediction Method of Fuzzy Association Rules in Interval Valued Relational Database

讨论了区间值关系数据库上模糊关联规则的挖掘算法与预测方法。采用一种比RFCM算法省时的FCMdd算法将记录在属性的取值划分成若干个模糊集,并提出区间值关系数据库上模糊关联规则的挖掘算法。仿真实例说明挖掘算法能够通过挖掘有意义的模糊关联规则来发现区间值关系数据库中蕴涵的关联性。区间值关系数据库上模糊关联规则的预测方法改进了标准可加性模型,并通过遗传算法调整模糊关联规则中三角模糊数的参数来提高预测的精度。

Mining algorithm and prediction method of fuzzy association rules are discussed in this paper.Fuzy relative of the k-medoids algorithm,which costs less time ,is adopted to partition values of the record in attribute into several fuzzy sets,and mining algorithm of fuzzy association rules in interval valued relational database is provided.Simulation example shows that this mining algorithm can discover the relationship contained in interval valued relational database by mining interesting fuzzy association rules.Prediction method of fuzzy association rules in interval valued relational database im-proves inference method of standard additive model,and genetic algorithm is applied to improve prediction accuracy by adjusting triangular fuzzy numbers in fuzzy association rules.