基于BP神经网络的河南省甲乙类法定报告传染病预测研究

发布时间：2019-06-02 07:37

【摘要】：目的建立用于河南省法定报告传染病(甲乙类)预测的神经网络模型,为制定传染病预防和控制措施提供理论依据。方法首先确定预测模型的基本结构,以归一化后的2003-2009年河南省甲乙类法定报告传染病发病率数据为训练样本,以2010年的数据为检验样本,采用改进的BP神经网络算法训练预测模型。利用该模型对2011-2013年河南省甲乙类法定报告传染病发病率数据进行预测。结果所建立的模型在仿真预测样本点的平均相对误差为0.076%,在检验样本处的预测误差为0.434%。并获得了2011-2013年河南省甲乙类法定报告传染病发病率预测数据。结论所建立的BP神经网络模型具有良好的预测精度,适合用来进行河南省甲乙类法定报告传染病发病率的预测。
[Abstract]:Objective to establish a neural network model for predicting legally reported infectious diseases (class A and B) in Henan Province, and to provide theoretical basis for the prevention and control of infectious diseases. Methods first of all, the basic structure of the prediction model was determined, and the standardized incidence data of category A and B legally reported infectious diseases in Henan Province from 2003 to 2009 were taken as training samples, and the data of 2010 were taken as test samples. The improved BP neural network algorithm is used to train the prediction model. The model was used to predict the incidence data of Class A and B legally reported infectious diseases in Henan Province from 2011 to 2013. Results the average relative error of the model at the simulation prediction sample point is 0.076%, and the prediction error at the test sample is 0.434%. The forecast data of the incidence of Class A and B legally reported infectious diseases in Henan Province from 2011 to 2013 were obtained. Conclusion the BP neural network model has good prediction accuracy and is suitable for predicting the incidence of Class A and B legally reported infectious diseases in Henan Province.
【作者单位】：河南中医学院基础医学院;
【基金】：河南省软科学研究重点项目(102400440002)
【分类号】：R183

【参考文献】