PERBANDINGAN DEFUZZIFIKASI CENTROID DAN MAXIMUM DEFUZZIFIER PADA METODE FUZZY INFERENCE SYSTEM UNTUK DIAGNOSIS PENYAKIT JANTUNG

Abstract

In the past decade, heart disease was ranked first of the highest cause of death in the world. As
revealed by the World Health Organization (WHO), as many as 7.4 million people died every year due
to this disease. Application of fuzzy logic is often used for several reasons, among them because of
the underlying mathematical concepts of fuzzy reasoning is very simple and easy to understand, very
flexible, has a tolerance of the data inappropriately. Basic fuzzy logic that is fuzzy set theory, in which
the role of determining the existence of degrees of membership as elements in a set is absolutely
essential. In addition the use of fuzzy logic in the current medical diagnosis is quite inflated. The main
processes in the Mamdani fuzzy logic starts from the input whose value was changed into fuzzy sets,
fuzzy inference, then defuzzification to return the fuzzy set into crisp output. The process of
defuzzification has an important role, because different defuzzification methods will produce the set
firmly. In this study compared defuzzifikasi centroid with maximum defuzzifier model mean of maxima.
The results for diagnosis using the defuzzifikasi centroid has the accuracy rate of 90.3% and for
defuzzifikasi mean of Maxima results of diagnosis is only 86,5%.