Expert System for Diagnosing Dengue Fever with Comparison of Naïve Bayes and Dempster Shafer Methods

An expert system for diagnosing dengue fever (DF) using a comparison of the Naive Bayes and Dempster Shafer methods aims to provide a solution to assist medical personnel in diagnosing this disease. Dengue fever is a disease caused by the dengue virus infection through the bite of Aedes mosquitoes. It has symptoms similar to other diseases and requires rapid and accurate diagnosis. The Naive Bayes and Dempster Shafer methods were chosen because both have different approaches to handling uncertainty and imprecise information. The Naive Bayes method is a probability-based classification that assumes independence between features. Meanwhile, Dempster Shafer is an approach to handling uncertainty. Therefore, comparing Naive Bayes and Dempster Shafer allows for classification with structured and fairly straightforward data, offering accuracy and flexibility in dealing with uncertainty. Applying this expert system with these methods can help in the faster and more accurate diagnosis of DF and provide better recommendations in situations where the available data is incomplete or ambiguous. From the test data calculations, the two methods show that the Naive Bayes method has a higher percentage value of 93%, while Dempster Shafer has 86%.