Application of seasonal autoregressive integrated moving average model in predicting incidence of infectious diarrhea in Fangshan district of Beijing

Objective To establish a seasonal autoregressive integrated moving average (SARIMA) model to predict the transmission trend of infectious diarrhea in Fangshan district of Beijing. Methods A SARIMA model was established based on the monthly incidence data of infectious diarrhea from 2004 to 2013 in Fangshan by using software R 3.0.1 TSA. We evaluated the fitting results of observed values and predicted values, and used this model to predict and analyze the transmission trend of infectious diarrhea by using the incidence data of infectious diarrhea in Fangshan from January to December 2014. Results SARIMA (0, 0, 2) (0, 1, 1)12 was fitted well with the observed values. The average relative error of the model fitted to the selected actual case data was 19.164%. The average relative error of the model in annual incidence was 2.303%. Conclusion SARIMA (0, 0, 2) (0, 1, 1)12 can be applied to predict short-term incidences of infectious diarrhea in Fangshan, which would provide scientific evidence for the evaluation of prevention and control of infectious diarrhea.