摘要
构建BP神经网络模型,对我国病毒性肝炎发病趋势进行预测.收集1990-2019年全国病毒性肝炎发病率的数据,分别构建ARIMA模型、支持向量回归、BP神经网络、ARIMA-SVR、ARIMA-BPANN模型,对病毒性肝炎发病率进行预测.实验结果表明,BP神经网络模型预测的效果最佳,得到测试集的拟合值和真实值的RMSE(平均残差平方和的平方根)和MAPE(平均绝对百分误差)分别只有0.427 3和0.385 8,远小于其他预测模型.
The trend of viral hepatitis in China was predicted by modeling.First,collect the relevant data of the incidence rate of viral hepatitis in China from 1990 to 2019,and build ARIMA model,support vector regression,BP neural network,ARIMA-SVR,ARIMA-BPANN and other models to predict the incidence rate of viral hepatitis.The experimental results show that the prediction effect of BP neural network is the best,The RMSE(square root of the mean residual sum of squares)and MAPE(mean absolute percentage error)of the fitting value and the real value of the test set obtained by this model are only 0.4273 and 0.3858,respectively,which are far less than other prediction models.
作者
徐香
朱家明
XU Xiang;ZHU Jiaming(The Fourth People's Hospital of Bengbu,Bengbu 233010,China;School of Statistics and Applied Mathematics,Anhui University of Finance and Economics,Bengbu 233030,China)
出处
《牡丹江师范学院学报（自然科学版）》
2023年第2期18-21,共4页
Journal of Mudanjiang Normal University：Natural Sciences Edition
基金
教育部人文社会科学研究项目(19YJCZH069)
安徽省教研项目(2020jyxm0017)。
关键词
病毒性肝炎
发病率
BP神经网络
预测
viral hepatitis
incidence rate
BP neural network
forecast
