Perbandingan Tingkat Akurasi Metode Fuzzy Mamdani dan Fuzzy Sugeno dalam Memprediksi Jumlah Kebutuhan Darah (Studi Kasus: PMI Kota Medan)

Abstract

Blood demand prediction is important to avoid instability in blood stocks or supplies. Too much and abundant blood stock can cause losses, due to the risk of wasted blood when it passes the storage limit period, while too little stock will risk a shortage of blood supply. To anticipate this, it is necessary to predict the amount of blood demand that must be received in the future in order to achieve optimum blood availability. Fuzzy Mamdani and Fuzzy Sugeno methods that use the concept of fuzzy logic are considered capable of handling complexity and uncertainty in decision making, so they can be used to predict the amount of blood demand by considering the varying and uncertain amounts of blood demand and supply. The calculation results show that the Fuzzy Mamdani method has better accuracy than the Fuzzy Sugeno method with MAPE values of 16.707% and 17.987% respectively. The Fuzzy Mamdani method has an accuracy level in the good category according to the MAPE method and can be used to predict blood demand at PMI Medan.