Li Jian, Gu Jun-zhong, Mao Sheng-hua, Xiao Wen-jia, Jin Hui-ming, Zheng Ya-xu, Wang Yong-ming, Hu Jia-yu
Shanghai Municipal Center for Disease Control and Prevention, Shanghai 200336, China.
Institute of Computer Application, East China Normal University.
Zhonghua Liu Xing Bing Xue Za Zhi. 2013 Dec;34(12):1198-202.
To establish BP artificial neural network predicting model regarding the daily cases of infectious diarrhea in Shanghai.
Data regarding both the incidence of infectious diarrhea from 2005 to 2008 in Shanghai and meteorological factors including temperature, relative humidity, rainfall, atmospheric pressure, duration of sunshine and wind speed within the same periods were collected and analyzed with the MatLab R2012b software. Meteorological factors that were correlated with infectious diarrhea were screened by Spearman correlation analysis. Principal component analysis (PCA) was used to remove the multi-colinearities between meteorological factors. Back-Propagation (BP) neural network was employed to establish related prediction models regarding the daily infectious diarrhea incidence, using artificial neural networks toolbox. The established models were evaluated through the fitting, predicting and forecasting processes.
Data from Spearman correlation analysis indicated that the incidence of infectious diarrhea had a highly positive correlation with factors as daily maximum temperature, minimum temperature, average temperature, minimum relative humidity and average relative humidity in the previous two days (P < 0.01), and a relatively high negative correlation with the daily average air pressure in the previous two days (P < 0.01). Factors as mean absolute error, root mean square error, correlation coefficient(r), and the coefficient of determination (r(2)) of BP neural network model were established under the input of 4 meteorological principal components, extracted by PCA and used for training and prediction. Then appeared to be 4.7811, 6.8921,0.7918,0.8418 and 5.8163, 7.8062,0.7202,0.8180, respectively. The rate on mean error regarding the predictive value to actual incidence in 2008 was 5.30% and the forecasting precision reached 95.63% .
Temperature and air pressure showed important impact on the incidence of infectious diarrhea. The BP neural network model had the advantages of low simulation forecasting errors and high forecasting hit rate that could ideally predict and forecast the effects on the incidence of infectious diarrhea.
建立上海市传染性腹泻日发病情况的BP人工神经网络预测模型。
收集上海市2005 - 2008年传染性腹泻发病率数据以及同期的温度、相对湿度、降雨量、气压、日照时长和风速等气象因素数据，运用MatLab R2012b软件进行分析。通过Spearman相关分析筛选出与传染性腹泻相关的气象因素。采用主成分分析（PCA）消除气象因素间的多重共线性。利用人工神经网络工具箱，采用反向传播（BP）神经网络建立传染性腹泻日发病率的相关预测模型。通过拟合、预测和预报过程对所建立的模型进行评估。
Spearman相关分析数据表明，传染性腹泻发病率与前两日的日最高气温、最低气温、平均气温、最小相对湿度和平均相对湿度呈高度正相关（P < 0.01），与前两日的日平均气压呈较高负相关（P < 0.01）。在输入经PCA提取并用于训练和预测的4个气象主成分的情况下，建立了BP神经网络模型的平均绝对误差、均方根误差、相关系数（r）和决定系数（r²）等指标，分别为4.7811、6.8921、0.7918、0.8418和5.8163、7.8062、0.7202、0.8180。2008年预测值与实际发病率的平均误差率为5.30%，预报精度达到95.63%。
温度和气压对传染性腹泻发病率有重要影响。BP神经网络模型具有模拟预测误差低、预报命中率高的优点，能够较好地预测和预报对传染性腹泻发病率的影响。
