Forecasting dengue fever cases using autoregressive distributed lag model with outlier factor

Abstract

Dengue is a mosquito-borne viral infection that has become one of the public health's biggest challenges. Dengue fever is transmitted through the Aedes Aegpty mosquito. One of the factors that can affect to dengue fever cases is climate factors. The climate change can make the number of dengue fever cases significantly increase. Therefore, the aims of this research are modeling and forecasting dengue fever cases by using Autoregressive Distributed Lag (ADL) model with outlier factor. ADL model is a model in time series which not only include time dependencies among itself, but also there are some other things that affect, which can be expressed as a predictor variable. Its model combines the present and past values of the dependent and independent variables. The observation that has different characteristics from other data is called outlier. The presence of an outlier has a certain meaning, which explanation can be lost if the outlier is removed. There are two special cases from types of outliers, Innovative Outlier (IO) and Additive Outlier (AO). The iterative procedure in detecting outliers in ADL model was introduced in this research. Therefore, the data containing outliers is neither deleted nor ignored, but still involves the outlier data by adding an outlier factor to the ADL model. The power of the procedure in detecting outliers are investigated by simulation experiments. The results are an ADL model with outlier factor that are free from outlier data and the predicted result of the number of dengue fever cases in the next two months.