【作者】 朱江华； 潘丰；

【Author】 Zhu Jianghua Pan Feng (Control Science and Engineering Research Center, Southern Yangtze University, Wuxi 214122, China)

【机构】 江南大学控制科学与工程研究中心；

【摘要】 给出了一种新的粗糙集知识约简方法,即结合粗糙集理论的依赖度定义,先给出知识约简转化定理,然后运用一种改进型蚁群算法,采用局部和全局搜索相结合的方法,对知识表达系统进行约简．同时,在适值函数中引入罚函数,从而保证所求的约简在包含最少而非零个属性的基础上有较大的依赖度．通过Matlab计算实例可看出,本文算法对求解知识约简问题快速有效．更多还原

【Abstract】 A new algorithm of rough set knowledge reduction is proposed. First a switch theorem about attribute reduction is given, and then an improved ant colony algorithm is proposed to solve the reduction of knowledge express system combined with decision attribute dependency degree. Meanwhile, in order to assure fewer attributes while not non-attribute, stronger decision attribute dependency degree in knowledge reduction, punishing function is used in fitness function. The practical results show that this approach is an effective and quick way in solving knowledge reduction.更多还原