基于小波神经网络的手足口病发病率预测研究-《集美大学学报（自然科学版）》

文章信息/Info

Title:: Predicting the Incidence of Hand-foot-mouth Disease Based on Wavelet Neural Network

Author(s):: QIU Xiuliang; XIAO Shixiao; ZHANG Baocan; (Chengyi University College,Jimei University,Xiamen 361021,China)

Keywords:: HFMD; meteorological factors; wavelet transform; wavelet neural network

摘要:: 结合气象因素，提出了基于小波神经网络的手足口病（hand-foot-mouth disease,HFMD）发病率预测模型。为验证模型有效性，进行了对比测试，测试结果如下：2014年7月至12月，厦门市手足口病实际月发病率为(7.4,4.7,24.3,21.1,8.2,2.8)×10－5，小波神经网络模型预测值为(4.3,21.3,15.9,3.5,5.1,27.2)×10－5，BP神经网络模型预测值为(20.1,14.3,1.7,10.6,68.2,0.4)×10－5，灰色预测模型预测值为(49.7,66.7,89.6,120.4,161.6,217.0)×10－5。通过实例分析表明，相比其他的传统的手足口病预测模型，小波神经网络模型具有收敛速度快、预测精度高、误差小的特点。

Abstract:: According to the research on the meteorological factors,a prediction model of incidence of HFMD was proposed based on wavelet neural network.In order to verify the validity of the model,a comparative test was carried out.The results were as follows:the actual monthly incidence of HFMD in Xiamen from July to December 2014 was (7.4,4.7,24.3,21.1,8.2,2.8)×10－5,and the predicted value of wavelet neural network was (4.3,21.3,15.9,3.5,5.1,27.2)×10－5,the predicted value of BP neural network was (20.1,14.3,1.7,10.6,68.2,0.4)×10－5,the predicted value of Grry predicition model was(49.7,66.7,89.6,120.4,161.6,217.0)×10－5.The results obtained from the simulator showed that the forecasting model of the wavelet neural network could evidently decrease prediction error and improve forecasting veracity compared with other models.