Portal de Periódicos da CAPES

Optimization of seasonal ARIMA models using differential evolution - simulated annealing (DESA) algorithm in forecasting dengue cases in Baguio City

2016; American Institute of Physics; Volume: 1776; Linguagem: Inglês

10.1063/1.4965385

1935-0465

Rizavel C. Addawe, Joel M. Addawe, Joselito C. Magadia,

Mosquito-borne diseases and control

Accurate forecasting of dengue cases would significantly improve epidemic prevention and control capabilities. This paper attempts to provide useful models in forecasting dengue epidemic specific to the young and adult population of Baguio City. To capture the seasonal variations in dengue incidence, this paper develops a robust modeling approach to identify and estimate seasonal autoregressive integrated moving average (SARIMA) models in the presence of additive outliers. Since the least squares estimators are not robust in the presence of outliers, we suggest a robust estimation based on winsorized and reweighted least squares estimators. A hybrid algorithm, Differential Evolution - Simulated Annealing (DESA), is used to identify and estimate the parameters of the optimal SARIMA model. The method is applied to the monthly reported dengue cases in Baguio City, Philippines.