摘要
针对BP神经网络的预测结果易受初始连接权值和阈值影响和陷入局部极值的问题,提出一种FOA优化BP神经网络的远程疾病诊断研究模型。在无线传感网络和ZigBee网络的基础上,通过FOA优化BP神经网络初始连接权值和阈值,实现BP神经网络初始连接权值和阈值的自适应最优选择。以威斯康幸大学医学院的乳腺癌数据集为研究对象,与PSO、GA和DE算法相比较,结果表明,FOA-BP可以有效提高疾病诊断的准确率。
Aiming at the problems that the prediction results of BP neural network are susceptible to the influence of initial connection weight and threshold value and fall into local extremum, a remote disease diagnosis research model of FOA optimized BP neural network is proposed. On the basis of wireless sensor network and ZigBee network, the initial connection weights and thresholds of BP neural network are optimized by FOA, and the adaptive optimal selection of initial connection weights and thresholds of BP neural network is realized. Using the breast cancer data set of School of Medicine, Wisconsin University as the research object, compared with the PSO, GA, and DE algorithms, the research results show that FOA-BP can effectively improve the accuracy of disease diagnosis.
作者
赵晓丹
ZHAO Xiaodan(Shanghai Fifth People's Hospital, Fudan University, Shanghai 200233)
出处
《微型电脑应用》
2019年第4期71-75,共5页
Microcomputer Applications
关键词
无线传感网络
ZIGBEE网络
果蝇算法
BP神经网络
疾病诊断
Wireless sensor network
ZigBee network
Fruit fiy optimization algorithm
BP neural network
Disease diagnosis