论文部分内容阅读
目的建立基于Elman神经网络的感染性腹泻病人中细菌性食源性疾病阳性检出率预测模型,评估探讨Elman神经网络模型在细菌性食源性疾病发病预测中的应用价值。方法利用深圳市2008年1月至2012年12月的细菌性食源性疾病疫情资料作为训练集,建立Elman神经网络模型;选取深圳市2013年1~6月的细菌性食源性疾病资料作为检验集,评价该模型的预测效能。结果当网络结构为12-32-1-1时,构建的Elman回归网络模型为最优预测模型,此时训练集模拟仿真结果的平均误差均方为65.75。在此最优网络预测模型下,检验集预测值的平均误差绝对值为1.20,平均误差绝对率为0.21,非线性相关系数为0.79。结论基于Elman回归网络预测模型对细菌性食源性疾病发病具有较好的预测效能。
OBJECTIVE: To establish a prediction model of the positive detection rate of bacterial foodborne diseases in patients with infectious diarrhea based on Elman neural network, and to evaluate the application value of Elman neural network model in predicting the incidence of bacterial foodborne diseases. Methods Using the data of epidemic situation of bacterial food-borne diseases from January 2008 to December 2012 in Shenzhen as a training set, an Elman neural network model was established. The data of bacterial foodborne diseases in Shenzhen from January to June 2013 were selected as Test set to evaluate the model’s predictive performance. Results When the network structure is 12-32-1-1, the Elman regression network model constructed is the optimal prediction model. The average error square of the training set simulation results is 65.75. Under the optimal network prediction model, the average absolute error of the test set prediction value is 1.20, the average absolute error rate is 0.21, and the nonlinear correlation coefficient is 0.79. Conclusion Based on the Elman regression network prediction model, it has a good predictive value for the onset of bacterial foodborne diseases.