Comparison of SARIMA model and Holt-Winters model in predicting the incidence of Sjögren's syndrome

Abstract

Objective: To analyze the prevalence trend of Sjögren's syndrome in the Department of Immunology and Rheumatology of Nanjing Zhongda Hospital from January 2015 to December 2019, and compare the application of SARIMA model and Holt-Winters model in predicting the number of cases of Sjögren's syndrome.

Methods: All of the data from the Department of Immunology and Rheumatology of Nanjing Zhongda Hospital were collected. The number of monthly cases from January 2015 to December 2019 was regarded as the training set, and it was used to establish the SARIMA model and Holt-Winters model. The number of monthly incidences from January 2020 to December 2020 was regarded as the test set, and it was used to check the model performance.

Results: The optimal model of SARIMA is ARIMA (0,1,1) (2,1,1)₁₂ model, and the optimal model of Holt-Winters model is Holt-Winters addition model. It was found that the Holt-Winters addition model produced the smallest error.

Conclusion: Holt-Winters addition model produces better prediction accuracy of the model.

Keywords: ARIMA model; Holt-Winters model; Sjögren's syndrome.