论文部分内容阅读
目的探索Elman神经网络在传染病预测中的应用价值。方法利用2005年1月～2010年6月我国内地法定报告的手足口病月发病率资料分别构建Elman神经网络、BP神经网络以及季节性自回归移动平均模型(SARIMA),并对3种模型进行预测效果评价。结果 Elman神经网络的预测结果的平均绝对误差(MAE)及均方误差平方根(RMSE)均小于BP神经网络、SARIMA模型。结论 Elman神经网络的预测效果较好,对于手足口病发病率预测具有较好的应用价值。
Objective To explore the value of Elman neural network in the prediction of infectious diseases. Methods Elman neural network, BP neural network and seasonal autoregressive moving average model (SARIMA) were respectively constructed from the statutory reported monthly incidence of hand-foot-mouth disease from January 2005 to June 2010 in our country. Three models Predictive effect evaluation. Results The mean absolute error (MAE) and square root of mean square error (RMSE) of Elman neural network were all less than those of BP neural network and SARIMA model. Conclusion The Elman neural network has a good predictive effect, which is of great value in predicting the incidence of hand-foot-mouth disease.