脑卒中发病率的统计分析及人工神经网络模型研究

Statistical Analysis and Artificial Neural Network Model of the Incidence of Cerebral Apoplexy

本文利用2005—2015年江西省万载县人民医院记录的脑卒中患者数据,通过SPSS17.0进行统计和Matlab 2012a软件构建Elman神经网络预测模型,得到了脑卒中与年龄、季节、职业之间都有紧密的关联,并通过拟合得出了各年脑卒中的预测数据,这些拟合预测数据的平均误差率(MER)在6.3%以下,而且拟合系数R2接近1,且准确率达到了90%以上。

By using the data of the cerebral apoplexy patients at Jiangxi Wanzai County People＇s hospital from 2005 to2015,the paper builds the Elman neural network forecasting model with MATLAB 2012 a Software after a statistical analysis on SPSS17.0. The results show that the disease is closely related to the age and profession of the patient,and the season. The analysis predicates the yearly data of the disease through fitting. The average error rate（MER）of the data is below 6.3%,the regression coefficient R2 is close to 1 and the accuracy rate more than 90%.