Forecasting Model to Predict the Incidence Number of Dengue Cases in Ubon Ratchatani Province, Thailand

Abstract

Abstract

This study aimed to determine the best model for forecasting the Dengue incidence number in the Ubon Ratchathani province. A comparison between Holt-Winter and Box-Jenkins technique was considered. This study collected data of all types of dengue fever from 2008 to 2018. In developing the model, a data size of 120 was used, whereas a data size of 12 was used for the validation.

The results showed that the ARIMA (0,1,0). SARIMA (0,1,0) was the best forecasting model of the Box and Jenkins technique. The MSE, MAPE, and RMSE values of the model development were 90.66, 252.95, and 127.77 respectively, and the validating step was 107.50, 54.72, and 169.36 respectively. But, it higher than the Holt-Winter technique which has the MSE, MAPE, and RMSE values of the model development were 53.88, 124.59, and 85.26 respectively, and the validating step was 107.50, 54.72, and 169.36 respectively. From the forecast of the number of dengue fever patients in 2019. We found that the tends to be epidemic is likely since the beginning of the year. The number of dengue patients was higher than the median value of the past 5 years' observations. The highest spread of the dengue fever was in May-August 2019.

In conclusion, the Holt-Winter was the best forecasting model to predict the Incidence Number of Dengue Cases in Ubon Ratchatani Province. However, the local public health officers should plan to monitor and prevent dengue fever and have intensive control of the disease in May-August 2019.

Keywords: Dengue fever, Model Holt-Winter, Model Box-Jenkins