基于遗传算法的粗糙集属性约简算法  

Rough Set Attribute Reduction Algorithm Based on GA

作　　者：李伟生[1] 易哲[1] 

机构地区：[1]重庆邮电大学计算机科学与技术学院,重庆400065

出　　处：《微电子学与计算机》2010年第3期71-74,共4页Microelectronics & Computer

基　　金：国家自然科学基金项目(60842003)

摘　　要：针对目前粗糙集遗传约简算法不能确保得到约简的不足,分别提出基于二进制编码基因和符号编码基因的遗传约简算法.基于二进制编码基因的遗传算法加入修正算子以确保遗传算法在可行解的空间搜索.基于符号编码基因的遗传算法融合交叉算子和变异算子以降低遗传算法的复杂度.实验结果表明,两种编码方式的遗传约简算法都能确保得到约简.To remove the drawbacks that the rough set genetic reduction algorithm can not essure to gain reductions, this paper presents two novel genetic reduction algorithms which are based on binary genes and character genes, respectively. The binary genes-based genetic algorithm is joined a modified operator to insure the algorithm to search in the possible solution space. The character genes-based genetic algorithm integrates the crossover operator and mutation operator for reducing the complexity of GA. The experiments results show that both of the genetic reduction algorithms can ensure reduction.

关 键 词：粗糙集 约简 遗传算法 二进制编码基因 符号编码基因 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]