Application Of ARIMA Model In Prediction Of Respiratory Infectious Disease

Objective To explore the application of time series autoregressive integrated movingaverage (Autoregressive integrated moving average, ARIMA) model for the predictionof respiratory infectious disease in this area. Provide decision-making basis formonitoring measures of preventing respiratory disease, while providing a reference foranother respiratory disease research.Methods Based on the data gathered in National Disease Reporting managementSystem, the prevalence situation of respiratory infectious diseases (measles, rubella,mumps, influenza, chicken pox, meningitis) in this district of Shushan in Hefei cityfrom2007to2011were studied. All statistical analyses were performed using thecomputer program SPSS version13.0. The ARIMA model is used to predict monthlyprevalence of the respiratory infectious diseases in2012, and the results are presented asshown. After parameter estimation, model diagnostics, model evaluation, we choose theoptimal model to predict the incidence of each month in2012.Results Two periods times from March to May and from November to January ofthe following year had the higher occurrence rate for6kinds of common respiratoryinfectious diseases in this district of Hefei city. After modeling and fitting, the ARIMA（0.0.1）model was determined to be the suited optimal model of common infectiousdisease, and RMSE was20.299, MAPE was41.264, the normalized BIC was6.226, thevalue of Ljung-BOX Q was0.375(p=0.375). The above fitting results showed that the residual error of the model accorded with the white noise standard. After modeling, wepredict the incidence of respiratory disease in our region in2012, the results show thatthe incidence of each month was similar to the model prediction. It show that theARIMA model can used to predict the trends and the incidence of those kinds ofrespiratory infectious diseases.Conclusions The forecasting results of the common infectious diseases in this districtbased on ARIMA model fitting was satisfactory, which could provide theoreticalsupport for common infectious diseases prevention and control of our area in the future.This research can provide theoretical support for the preventing and controlling ofcommon respiratory infectious diseases. Choosing the optimal ARIMA model toforecast and warn when respiratory infectious diseases is coming, it has a practicalvalue for strengthening the routine of prevention and control of respiratory infectiousdiseases.