摘要
首先分析了专家知识库粗集建模中连续数据离散化存在的问题 ,指出允许引入少量的冲突对专家知识库的建模分析是有益的 ;提出了一种基于信息熵的数据离散化方法 ,并分析了数据离散化的熵的度量 ,根据求解问题设计了一种问题求解的遗传算法 ;最后以基于多 Agent车间调度系统中调度
The paper analyze the problems of th e real-value disretication in a rough set model for a expert knowledge base and point out that the permission to introduce a few data conflicts in the process o f modeling is helpful to the analysis of expert knowledge at first. A method for dicretization of attribute values based on the information entropy is presented , in which we measure the dicretization with the entropy gain of an attribute an d we construct a genetic algorithm for the method. Finally we give an example of the application of the method that is the disretication in modeling of the know ledge base of scheduler agents.
出处
《数学的实践与认识》
CSCD
北大核心
2003年第8期60-65,共6页
Mathematics in Practice and Theory
基金
国家 8 63 /CIMS项目 (863 -5 11-910 -4 0 3 )
江苏省教育厅自然科学基金重点项目 (0 1KJB5 2 0 0 10 )
