摘要
目的建立武汉市洪山区的学生肺结核预测模型,提供科学精准的结核病防控策略。方法根据2013—2020年学生结核病月病例数建立时间序列模型,选取最优模型,并进行模型预测效果评价。结果拟合度和预测效果较好的最优模型为ARIMA(0,0,3)(0,1,1)_(12),所得到的预测值与实际值趋势基本一致,且实际值均落在预测值的95%可信区间内,MAPE=2.313%,模型的预测效果较好。结论ARIMA乘积季节模型在洪山区学生肺结核发病时间序列中能很好地拟合和预测,可以用于洪山区学校肺结核发病预警预测。
Objective To establish early forecasting model of tuberculosis in schools of Hongshan district,and provide scientific strategy for prevention and control of tuberculosis.Methods Using data on 2013-2020 in schools tuberculosis,established time series models and chosen the best one to forecast and tested the incidence of tuberculosis.Results The best model for school tuberculosis was ARIMA(0,0,3)(0,1,1)_(12).The trend of the predicted value was basically consistent with the actual value which was also in the 95%confidence interval of predicted numbers.The effectiveness of forecasting was good with a 2.313%value of MAPE.Conclusion The ARIMA product seasonal model can effectively fit and forecast time series data on students′tuberculosis.It also can be used to early warn and predict the incidence of school tuberculosis in Hongshan district.
作者
李丽
刘亦新
杨云
刘静
王丽娜
LI Li;LIU Yi-xin;YANG Yun;LIU Jing;WANG Li-na(Hongshan District Center for Disease Control and Prevention,Wuhan,Hubei 430064,China;School of Computer and Communication Engineering,Zhengzhou University of Light Industry,Zhengzhou,Henan 450001,China)
出处
《公共卫生与预防医学》
2022年第5期106-110,共5页
Journal of Public Health and Preventive Medicine
基金
河南省自然科学基金(212300410303)。
关键词
学生
结核病
时间序列
预测
Students
Tuberculosis
Time series
Forecasting
