Application of autoregressive integrated moving average model in prediction of other infectious diarrhea in Liaoning province

  Objective  Through the analysis on the monthly incidence of other infectious diarrhea in Liaoning province from 2007 to 2017, an autoregressive integrated moving average (ARIMA) model was established to provide reference for the prevention and control of other infectious diarrhea in Liaoning.

  Methods  The monthly incidence data of other infectious diarrhea in Liaoning from 2007 to 2016 were collected from the Public Health Science Data Center of the National Population and Health Science Data Center for an analysis with software. SPSS 25.0, an ARIMA model based on the time series was established by using the incidence rate to predict the monthly incidence rate of other infectious diarrhea in Liaoning in 2017, and the accuracy of the model prediction was evaluated based on the actual value.

  Results  The time series of the incidence of other infectious diarrhea in Liaoning from 2007 to 2016 was a non-stationary one. Based on graphical observation and multiple verifications, the following four alternative models were determined: ARIMA (1,1,1) (0,1, 0)₁₂, ARIMA (1,1,1) (1,1,0)₁₂, ARIMA (1,1,1) (0,1,1)₁₂ and ARIMA (1,1,1) (1,1, 1)₁₂. Through the Ljung-Box test and comparing Bayesian Information Criterion (BIC) value, ARIMA (1,1,1) (0,1,1)₁₂ was finally determined to be the optimal model. Compared with the actual value of each month in 2017, the model prediction was highly accurate.

  Conclusion  The ARIMA (1,1,1) (0,1,1)₁₂ model can better predict the monthly incidence of other infectious diarrhea in Liaoning. It is necessary to promote the application of this model.