基于遗传算法的肝病诊断学习系统  

A Learning System Based on Genetic Algorithms for Liver Disease Diagnosis

作　　者：王遵亮[1] 吴新根[1] 罗立民[1] 

机构地区：[1]东南大学生物科学与医学工程系

出　　处：《东南大学学报（自然科学版）》1999年第3期106-109,共4页Journal of Southeast University：Natural Science Edition

摘　　要：提出一种用遗传算法来构造肝病诊断学习系统的方法，其中引入了肝病领域专家对规则打分的适应度评估机制，通过选择、交叉和交异等遗传操作使知识规则集得到不断地精炼．最后，用这种方法对模拟的专家知识进行学习，实验结果表明，该方法能使系统比较有效地学习专家知识．A method which is used to construct a learning system for liver disease diagnosis using genetic algorithms is presented. A new fitness function in which adopted the mechanism that the domain expert evaluates the rules is provided in this method. Through the use of genetic operators reproduction, crossover and mutation, the rules is gradually refined. Finally, we applied the method in the learning system to learn the experts' knowledge. The experimental result shows our method can make the learning system to learn domain experts' knowledge efficiently.

关 键 词：遗传算法 学习系统 专家系统 肝病 诊断 

分 类 号：R575.04[医药卫生—消化系统] R319[医药卫生—内科学]