Using R To Fit A Hybrid ARIMA And Support Vector Machines Model In Tuberculosis Incidence Forecasting Of Jiangxi Province

Objective: To arrange the epidemic data of the monthly incidence of tuberculosis in Jiangxi Province, from 2004 to 2012. To explore the application of the single auto regressive integrated moving average model(ARIMA) and the auto regressive integrated moving average-support vector machine(ARIMA-SVM) hybrid model which may be used as a decision-supportive tool for planning health interventions and allocating health resources in the fitting and forecasting research of the monthly incidence of tuberculosis. Methods:An example of monthly tuberculosis incidence in Jiangxi province from 2004 to 2012 was given to fit the ARIMA-SVM hybrid model to predict the incidence using forecast package and e1071 package in R. The fitting and prediction accuracy of the two models was evaluated by mean squar eerror(MSE) and mean absolute percentage error(MAPE). Results:The time series of tuberculosis report rate show a gradual secular decline and a striking seasonal variation. The ARIMA(0,1,1)(2,0,0)12 model was selected from several plausible ARIMA models. The parameter of SVM model were radial function, C=100, γ=0.01, ε=0.1. The MSE of ARIMA and ARIMA-SVM model were 5.044 and 4.837, respectively. The MAPE of ARIMA and ARIMA-SVM model were 0.318 and 0.316, respectively. Conclusions:The ARIMA-SVM hybrid model showed better tuberculosis incidence prediction than the ARIMA model. The seasonality of the tuberculosis incidence suggested by the models indicate the need for interventions focused on reducing infectious dieease transmission on public transportion during the Sping period. We should direct programmatic interventions and allocate public health resources continuously since the a gradual decline of tuberculosis incidence showed.