气象因素和百度指数在手足口病预报模型中的应用

Application of meteorological factors and Baidu index in hand-foot-mouth disease prediction model

目的 结合手足口病关键字的百度指数与手足口病病例监测数据,建立手足口病预测模型。方法 基于气象因素和百度指数,利用上海市浦东地区2012年1月1日-2016年12月31日手足口病发病数据,构建3种预测模型,使用解释偏差和赤池信息准则比较3种模型的拟合优度。利用交叉验证的方法,通过计算残差平方和和调整后的R^(2),反映预测值与实际观察值之间的差异情况、评估预报模型的性能。结果 2012-2016年,上海市浦东地区共报告手足口病病例63 005例,平均每周(241.39±170.95)例,对比3种模型的拟合结果,模型2的拟合效果最佳,其手足口病病例的预测指标优于模型3和模型4的。结论 与气温和相对湿度相比,百度指数的预测时效性更强,利用气象数据和百度指数可以提高手足口病预测模型的预测能力。

Objective The combination of Baidu index of hand-foot-mouth disease(HFMD)keywords and monitoring data of HFMD case was used to explore and establish the optimal prediction model of HFMD.Methods The incidence data of HFMD in Pudong New Area of Shanghai from January 1,2012 to December 31,2016 were collected.Three prediction models are constructed based on meteorological factors and Baidu index.The actual observed number of HFMD patients per week is recorded as the observed value,the number of HFMD patients per week obtained by the prediction model is recorded as the predicted value.The model is established on the training set,the goodness of fit of the three models is compared by using interpretation deviation and Akaike's information criterion.Results A total of 63005 HFMD cases were reported in Pudong New Area of Shanghai from 2012 to 2016,with an average of 241.39±170.95 cases per week.Comparing the fitting degree of the three models,model 2 has the best fitting effect.Model 2 is a better predictor of HFMD cases,compared with model 3 and model 4.Conclusion Compared with temperature and relative humidity,Baidu index has stronger timeliness.The performance of HFMD prediction model can be improved by using meteorological data and Baidu index.