Abstract
COVID-19 is an infectious disease caused by a coronavirus which spreads from direct human contact through droplets of mucus in the respiratory tract of an infected person.The American Centers for Disease Control and Prevention (CDC) says that asymptomatic COVID-19 patients may account for more than 50% of the transmission rate.This research uses the SVM (Support Vector Machine) model as a feature extraction processor from voice data in the training and testing process, so that it can detect asymptomatic COVID-19 from the extraction of cough voice recordings.Of the 171 subjects studied, 120 subjects (70%) for training data and 51 (30%) for test data.The data is divided into the SMOTE data and without the SMOTE data process.The results of the two data have an average performance matrix of above 80%, with accuracy for without the SMOTE data of 98.3% and for SMOTE data of 100%.