正态云关联规则的预测方法

PREDICTION METHOD OF NORMAL CLOUD ASSOCIATION RULES

在信息处理领域,用数据挖掘方法发现关联规则和进行预测是两大热点。文中应用聚类的方法确定正态云的两个参数,并借助正态云模型来划分数量属性的论域,由此生成一系列的正态云关联规则。接着给出了正态云关联规则的挖掘和预测方法。由于用正态云表示的语言值能很好地表达抽象的概念,从而使得挖掘出的正态云关联规则与预测的结果更抽象、更容易被人理解。

Mining association rules and making prediction by data mining method are two hotspots in the field of information processing. In this paper, clustering method is adopted to determine two parameters of a normal cloud, then the normal cloud model is adopted to partition the domains of the quantitative attributes and a series of normal cloud association rules are generated. The mining and prediction methods of the normal cloud association rules are also provided in this paper. Because the abstract concepts can be well expressed by the linguistic values that are expressed with the normal clouds, the normal cloud association rules mined and the prediction results are more abstract and understood easily by people.