Evolution of predictive model for Dengue incidence by using machine learning algorithms

Dengue is a mosquito-borne, fatal viral disease, which is recently expanding as a global problem. There are three main crucial climatic factors namely temperature, rainfall, and humidity remains critical to the mosquito survival, reproduction, and development, which can further influence the mosquito presence and abundance. A Predictive model for the dengue incidence and its prevalence has been proposed by considering Kerala state's Dengue data. In prediction module the algorithm predicts the number of dengue cases that might occur in that month. In meteorological analysis the relationship between humidity, temperature, rainfall and dengue cases were studied. The machine learning algorithms employed for the prediction were Linear Regression, SVR and Kernel Ridge. Comparative study has been done on meteorological and geo spatial analysis on the data. The parameters taken into consideration for geo-spatial analysis were district's distance from the equator, shortest distance from the nearest seashore & percentage of forest cover in district, and then the altitude of that particular district.