Journal of Public Health and Preventive Medicine 2020;31(6):29-32

doi:10.3969/j.issn.1006-2483.2020.06.007

Comparative study of ARIMA model and seasonal index model in the prediction of mumps in Hubei Province

Peng ZHANG 1 ; Jing CAI 1 ; Shuqiong HUANG 1 ; Wenwen YANG 1 ; Cong XIE 1 ; Ran WU 1

Affiliations

+expand

Keywords

ARIMA model; Seasonal index model; Mumps; Prediction

Country

China

Language

Chinese

Abstract

Objective To establish an ARIMA model and a seasonal index model to predict the trend of mumps, compare the advantages and disadvantages of the two models, and to provide a scientific basis for the prevention and control of mumps. Methods ARIMA model and seasonal index model were established based on the monthly incidence of mumps in Hubei Province from 2008 to 2019. Results The average annual incidence rate from 2008 to 2019 was 28.89 / 100,000. April-July was the month of high incidence. The established ARIMA model and seasonal index model were （1-1.070B+0.441B2-0.291B3）*（1-B12）*Xt=（1-0.611B12）*Ɛt and Xt=（2.802-0.006t）*St. The average relative errors of the ARIMA model and the seasonal index model were 11.49% and 20.86%, respectively. Conclusion The ARIMA model and the seasonal index model both have good applicability in predicting the onset time characteristics and trend of mumps. However, while the ARIMA model demonstrated more advantages in fitting the annual change trend, the seasonal index model is better in fitting the monthly change trend. The two models can be used in combination to predict the trend of mumps.