Klasifikasi pendonor darah potensial menggunakan pendekatan algoritmepembelajaran mesin

Abstract – Blood donation is the process of takingblood from someone used for blood transfusions.Blood type, sex, age, blood pressure, and hemoglobinare blood donor criteria that must be met andprocessed manually to classify blood donor eligibility.The manual process resulted in an irregular bloodsupply because blood donor candidates did not meetthe criteria. This study implements machine learningalgorithms includes kNN, naïve Bayes, and neuralnetwork methods to determine the eligibility of blooddonors. This study used 600 training data divided intotwo classes, namely potential and non-potentialdonors. The test results show that the accuracy of theneural network is 84.3 %, higher than kNN and naïveBayes, respectively of 75 % and 84.17 %. It indicatesthat the neural network method outperformscomparing with kNN and naïve Bayes