Dengue Fever has existed throughout the contemporary history of mankind and poses an endemic threat to most tropical regions. Dengue virus is transmitted to humans mainly by the Aedes aegypti mosquito. It has been observed that there are significantly more Aedes aegypti mosquitoes present in tropical areas than in other climate regions. As such, it is commonly believed that the tropical climate suits the life-cycle of the mosquito. Thus, studying the correlation between the climatic factors and trend of dengue cases is helpful in
conceptualising a more effective pre-emptive control measure towards dengue outbreaks. In this chapter, a novel methodology for forecasting the number of dengue cases based on climactic factors is presented. We proposed to use Wavelet transformation for data preprocessing before employing a Support Vector Machines (SVM)-based Genetic Algorithm to select the most important features. After which, regression based on SVM was used to perform forecasting of the model. The results drawn from this model based on dengue data in Singapore showed improvement in prediction performance of dengue cases ahead. It has also been demonstrated that in this model, prior climatic knowledge of 5 years is sufficient to produce satisfactory prediction results for up to 2 years. This model can help the health control agency to improve its strategic planning for disease control to combat dengue outbreak. The experimental result arising from this model also suggests strong correlation between the monsoon seasonality and dengue virus transmission. It also confirms previous work that showed mean temperature and monthly seasonality contribute minimally to
outbreaks.