摘要
为了提高癌细胞识别正确率,弥补机械地以划定数值界限做出判决的缺陷,提出了利用直觉模糊集的概念和方法,并结合医生的先验细胞识别知识,进行癌细胞识别.建立了细胞样本集到症状集,症状集到诊断分类集上的直觉模糊关系.根据合成运算,得到了判断细胞样本所属诊断分类集上的肯定,否定,犹豫程度.最后通过一个示例的学习过程表明,方法能够达到正确识别分类,是合理有效的.并且可以根据医生新的实践经验做出相应的调整,此方法可进行不断地学习,直至达到满意的结果.
In order to improve the diagnostic accuracy of cancer cells and to make up for the defects of the judgment by the strict numerical value limits, a new cancer cells recognition method was put forward base on intuitionistic fuzzy sets and doctors＇ prior knowledge. The intuitionistic fuzzy relations were established from a cell sample set to a set of symptoms and from a symptoms set to a classification set of cells diagnosis. The evaluation intuitionistic fuzzy sets on the cell sample was obtained with the composition operation of the intuitionistic fuzzy relations. Finally, an example-based learning approach was presented. The study result indicates that the new method can correct to classify and identify cells, which is reasonable and effective. In addition, this recognition method can continuously renovate and perfect by doctors＇ new practical experience in the application to advance precision.
出处
《数学的实践与认识》
CSCD
北大核心
2013年第16期112-117,共6页
Mathematics in Practice and Theory
基金
四川省教育厅基金(11ZB153
10ZC024
13ZA0139)
国家自然科学基金(11071178)
四川省教育厅青年基金(10ZB027)
四川民族学院科研基金(12XYZB006)对本文给予了资助
关键词
模糊集
直觉模糊集
关系合成
fuzzy sets
intuitionstic fuzzy sets
relation composition
