A Data Mining Framework for Building Dengue Infection Disease Model

Dengue infection is a virus-caused diseased that is spread by mosquitoes. Currently, there is no specific treatment, moreover there is no vaccine to prevent the people from the disease. Symptoms of this infection show rapid and violent to patients in a short time. Two classification problems are explored in this study, which are the Dengue Classification Problem and the Day of defervescence of fever (day0) detection problem. Two data mining models, fuzzy logic and decision tree, are studied in this work. We propose to use the knowledge obtained from the decision tree with the fuzzy logic approach in order to obtain better performance. The experimental result shows that using fuzzy logic approach for Dengue Classification is a suitable method. We get 97.39% of the average accuracy from fuzzy logic approach. Consider the day0 detection problem, using the attributes obtained from the dengue patient can provide the early warning on one day before the day0 with the sensitivity as 71.11% by using fuzzy logic.