摘要
目的:运用人工神经网络(ANN)模型,对四种疾病进行辅助诊断。方法:从100例四种疾病的病例中随机抽取70例作为训练集,15例作为验证集,15例作为测试集,通过对网络进行训练,用训练好的网络对验证集和测试集进行仿真诊断试验。结果:ANN诊断消化道系统疾病的准确率达到93.7%以上。结论:ANN模型基本上可以用来在临床上作辅助诊断。
Purpose： The use of artificial neural network （ANN） model for the auxiliary diagnosis of the four diseases. Method： From 100 cases of the four kinds of diseases were randomly selected, 70 cases as the training set, 15 cases as a validation set, and 15 cases as a test set. After the training of the network, using the trained network on validation set and test set to simulate diagnostic test. Result： The accurate rate of ANN diagnosis of digestive tract diseases reached more than 93.7%. Conclusion： ANN model can basically be used in clinical diagnosis.
出处
《科技广场》
2016年第7期162-164,共3页
Science Mosaic
基金
2014年大学生创新创业教育项目
2014江西中医药大学重点学科青年教学培养项目(编号:jzzdxk018)
关键词
神经网络
训练函数
学习函数
传递函数
Neural Network
Training Function
Learning Function
Transfer Function
