Abstract:

BACKGROUND: The incidence rate of pulmonary heart disease is affected by many factors, and each kind of factor is maintaining the intriguing relation with others. Therefore, it is difficult to analyze and predict using deterministic mathematical model.
OBJECTIVE: To investigate the application of ARMA forecasting model on time series data, and to establish forecasting model on pulmonary heart disease incidence rate in Haixizhou region.
METHODS: The data of pulmonary heart disease incidence rate in Haixizhou from January 2003 to December 2008 were analyzed by time sequence. Three kinds of ARIMA models had been established through data stabilizing and mode identifying, and ARIMA (2, 1, 1) model has been selected out by virtue of the criterion of AIC and SC. At last, the rationality of the model was ensured by model test and model predication.
RESULTS AND CONCLUSION: The dynamic trend of prediction value forecasted by ARIMA (2, 1, 1) model basically accorded with the actual condition, with quite ideal results. ARIMA (2, 1, 1) model can be used as the forecasting of pulmonary heart disease incidence rate, which can help people comprehend the variation trend and regularity for seasonal change of pulmonary heart disease incidence rate, focused on the work of pulmonary heart disease healthy protection, effectively reduce the hazards of pulmonary heart disease to human, protection of human life quality.