利用粗糙集理论可以从已知数据中挖掘决策规则 .对于连续取值的特征属性必须先对其离散化 .从给定的特征属性集合中去除冗余的特征属性 ,选取有用的属性子集有助于简化决策规则 .我们利用基于信息熵的规则不确定性量度函数构造了一个决策规则挖掘的遗传算法 ,将规则挖掘与特征选取和连续属性的离散化集成在一起 .实验结果说明了这种方法的有效性 Rough set theory can be used to mine decision rules from known data, which must be discretized firstly for the continually valued feature attributes, remove the redundant feature attributes from the given set of feature attributes, select the useful attribute subset Which can help to simplify decision rules.We construct a genetic algorithm for decision rule mining by using the rule-based uncertainty measure function based on information entropy, and integrate the rule mining with feature selection and discretization of continuous attributes.The experimental results show that this The effectiveness of the method