Time series analysis of dengue cases in Central Luzon: comparison of stochastic models using climate variables as predictors

This is the authors' abstract. We don't add key points for this paper.

This study compares multiple stochastic models generated in order to identify which of them best fits the data. Monthly dengue cases from 2003 to 2013 in Central Luzon were used to generate the model. Seasonal Autoregressive Integrated Moving Average (SARIMA) is used to predict dengue cases (RMSE = 0.442) based on past dengue outbreaks with climate variables as predictors. It shows that the predictive variables, individually, brings about significant changes with the predicted dengue occurrence, lowering its RMSE to 0.402, and 0.403. Via cross correlation, it is found that the minimum temperature has the highest correlation to dengue occurrence (r = 0.6332) with RMSE = 0.402. The root mean square error is then used to compare and determine which among the stochastic models best forecasts dengue incidence.