粗糙集理论的主要应用是属性约简和规则提取,但由于应用粗糙集理论提取出的规则未必都是最佳规则,因此,本文提出一种基于互信息的规则约简方法。对确定性规则进行优化,挖掘出最简规则集,最后通过实例分析验证了该方法可行性和有效性。 The main applications of rough set theory are attribute reduction and rule extraction. However, since the rules extracted by applying rough set theory are not always the best rules, we propose a rule reduction method based on mutual information. The deterministic rules are optimized and the simplest rule sets are extracted. Finally, the feasibility and validity of the method are verified by an example analysis.