摘要
选用隶属度函数为高斯函数的模糊神经网络用于肺癌诊断,尝试提高诊断的正确率.对于非二值输入参数,首先用高斯隶属度函数模糊化,然后与二值参数一起作为BP神经网络的输入参数.所用病例被随机分为训练集和证实集,训练模糊神经网络,用证实集测试该网络区分肺癌与非肺癌的能力.结果表明,用高斯隶属度函数的模糊神经网络比作为对照的三角形隶属度函数模糊神经网络诊断正确率有所提高,而且对病例如何分组不敏感.
The usefulness of a fuzzy neural network（FNN） with Gaussian membership function（GMF） for distinguishing between lung cancer and benign cases was studied to improve lung cancer diagnosis.Thirteen non-binary parameters were fuzzed using GMF.The fuzzed outputs added with the other 13 binary parameters served as inputs of the BP neural network.Including lung cancer and benign cases,117 cases were used to train the FNN.Tens of cases were sampled from the 117 cases at random as training set,and the other cases as validation set.The validation set was used to test the performance of the trained FNN.The performance of the FNN with GMF was compared with that of the FNN with triangle membership function（TMF）.The performance of the FNN with GMF was better than that of the FNN with TMF.
出处
《郑州大学学报（理学版）》
CAS
北大核心
2011年第1期95-98,共4页
Journal of Zhengzhou University:Natural Science Edition
基金
国家自然科学基金资助项目
编号30571552
关键词
高斯隶属度函数
模糊神经网络
三角形隶属度函数
肺癌诊断
Gauss membership function
fuzzy neural network
triangle membership function
lung cancer diagnosis
