基于遗传编程的分等级规则挖掘模型

根据分等级规则挖掘的具体要求，结合遗传编程的特点建立了一种新型的规则挖掘模型。对规则的编码方式和遗传算子进行了详细的设计，构造了满足遗传编程挖掘的树型染色体规则结构，并综合考虑支持度、可信度和相关度3个指标和等级系数等约束因素来设计适应度函数。最后通过实例对基于遗传编程的分等级规则挖掘模型的可靠性进行了验证。

Hierarchical rules mining model based on genetic programming

According to the requirements of mining hierarchical rules and the characteristic of the genetic programming(GP),a hierarchical rule mining pattern based on GP has been proposed.To construct the chromosome tree of the hierarchical rules that is necessary for GP,the coding pattern and genetic operators of the hierarchical rule mining based on GP has been designed.To design the fitness function,three indicators-support,confidence,and correlativity,and some other constraints such as hierarchical coefficients were considered.Finally,a case has been employed to demonstrate the reliability of the GP-based hierarchical rules mining model.