Using adaptive fuzzy systems for controlling dengue epidemic in Sri Lanka

Dengue epidemic is one the hard challenges that Sri Lankan citizen face today. With the fast growth and due to unavailability of medicines, situation has been worsened. The only way to thwart this danger is to extinct the main cause Aedes aegypti mosquito. Current activities to minimize the mosquito population, are done in an ad-hoc manner. This paper proposes a methodology to recognize the patterns of mosquito spread to increase the effectiveness of the national dengue controlling program. Many climate and socio-economic factors such as temperature, precipitation and urbanization are correlated with the dengue spread. By providing those parameters as inputs and records of reported dengue cases as training data to an adaptive fuzzy system, vulnerability of a particular location to dengue can be obtained as the output. Output will estimate ´how dengue is high´ as a fuzzy value between 0 and 1. The solution is based on adaptive neuro fuzzy systems and k-means clustering.