| Summary: | Dengue is the most important arthropod-borne viral disease of humans which the global dengue incidence has grown dramatically in recent decades. Annual dengue incidence reported in the past five decades in 1955 to 2010 has increased thirty- fold. Currently, the number of dengue cases reported increased over eight-fold within the last two decades which the spike in the number of dengue cases globally was in 2019 compared to 505,430 cases in 2000. This epidemic dengue became a major public health problem in several countries including Malaysia with many regions under conditions of rainy season with warmer temperatures and high humidity allowing vectors to flourish and create access for dengue transmission. This study aims to develop a potential models and compare the best models based on the epidemiological week (epi-week) dengue data in Selangor, Malaysia focusing on Gombak and Klang districts. The Box-Jenkins method applies ARIMA model to find the best fit of dengue cases from 2013 to 2019 for both districts. The estimation part known as a model development for dengue data 2013 to 2017, while the evaluation part adopted dengue data from 2018 to 2019. The estimation period was evaluated by using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC), whereas error measures to select the best models in evaluation period used Mean Square Error (MSE), Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). The model that satisfied all criterion have been chosen as the best model in the evaluation part. The result revealed that the best model that closely described the existence of epi-week dengue cases for Gombak and Klang districts are ARIMA (1,1,1) and ARIMA (3,1,2) respectively. © 2023 Author(s).
|
|---|