Abstract:Objective To explore the feasibility of application of seasonal auto-regressive integrated moving average (SARIMA) model in predicting the monthly number of bacillary dysentery in Suzhou City. Methods R i386 3.2.3 software was used to establish SARIMA model based on the data regarding the monthly number of bacillary dysentery in Suzhou City from January 2005 to April 2018. The monthly number of cases of bacillary dysentery in Suzhou City from May to July in 2018 was forecasted, and the prediction effect was evaluated. Results The model of SARIMA (0, l, 2)×(0, l, 1)12 was established. Ljung-Box test showed that the prediction results with the model accorded well with the actual data (Q=19.194, P=0.244), all the actual values of monthly number of cases of bacillary dysentery in Suzhou City from May to July in 2018 fell in the 95% confidence intervals of expected values, and the mean relative error was -0.147. Conclusion The SARIMA (0, l, 2)×(0, l, 1)12 model is well fit for predicting the monthly number of cases of bacillary dysentery in Suzhou City.
王建书, 刘强, 覃江纯, 杭惠, 杨海兵. SARIMA模型在苏州市细菌性痢疾发病预测中的应用[J]. 实用预防医学, 2019, 26(6): 656-658.
WANG Jian-shu, LIU Qiang, QIN Jiang-chun, HANG Hui, YANG Hai-bing. Application of seasonal auto-regressive integrated moving average model to predicting the incidence of bacillary dysentery in Suzhou City. , 2019, 26(6): 656-658.