Abstract：Objective To apply the autoregressive integrated moving average （ARIMA） model to predicting the tuberculosis （TB） incidence. Methods Based on the data of TB incidence in Yuyao from 2006 to 2016，the ARIMA model was established using the Expert Modeler and the traditional modeling process. The optimal model was selected according to the minimum value of Bayesian information criterion （BIC） to fit the monthly incidence rate of TB from 2006 to 2016 and to predict the incidence of TB in 2017. Results The traditional modeling process established ARIMA（0，1，1） （0，1，1）12 and the Expert Modeler established ARIMA（0，0，1） （0，1，1） 12 . The two models' residual sequences did not break through the confidence intervals and both were appropriate. The ARIMA （0，0，1） （0，1，1） 12 was the optimal model due to a smaller standardized BIC value. When fitting the monthly incidence of TB in Yuyao from 2006 to 2016，the actual incidence of TB fell into 95% confidence intervals of the fitting value and the predicted value could fit the original data. When predicting the monthly incidence of TB in 2017，the mean relative error was 9.05%. Conclusio The ARIMA model constructed by Expert Modeler is suitable for predicting TB incidence.
胡碧波，傅克本，许亮亮，何丽萍. 应用ARIMA模型预测结核病发病率研究[J]. 预防医学, 2018, 30(10): 1011-1015.
HU Bi-bo，FU Ke-ben，XU Liang-liang，HE Li-ping. Application of ARIMA model to prediction of tuberculosis incidence. Preventive Medicine, 2018, 30(10): 1011-1015.
AGARWAL S K，SINGH A，ANURADHA S，et a1.Cytokine profile in human immunodeficiency virus positive patients with and without tuberculosis[J] . The Journal of the Association of Physicians of India，2001，49：799-802.
LI Q，GUO N N，HAN Z Y，et al. Application of an autoregressive integrated moving average model for predicting the incidence of hemorrhagic fever with renal syndrome[J] . Am J Trop Med Hyg，2012，87（2）：364-370.