摘要
为解决冠心病多参数诊断中存在的问题,提出了一种基于优化算法的神经网络智能诊断系统,即采用优于BP算法的LM算法进行网络学习训练,采用黄金分割优选法对网络隐含层节点数进行优选。临床实验应用表明,这种系统具有拟合精度高和运算速度快等优点,且诊断正确率为100%。
To solve the problems of multi-parameter coronary heart disease diagnosis,a neural network based on optimization algorithm is presented,using a better Levenberg-Marquardt algorithm compared to BP algorithm and the method of golden section applied on the optimization of the hidden layer nodes.The application of clinic experimentation shows the diagnosis system possesses the merits of high precision in sample fitting and calculation speed,and the ratio of correctness of diagnosis is 100%.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第35期197-199,共3页
Computer Engineering and Applications
基金
国家自然科学基金项目(编号:60173058
60377020)资助
关键词
神经网络
智能系统
冠心病诊断
neural network,intelligent system,coronary heart disease diagnosis