Machine learning to detect covid-19 using cough sounds

Abstract

The objective of this project is to design a diagnostic aid system for the early COVID- 19 Detection disease from the voice. Mainly, the proposed system is based on two main steps: feature extraction of sound and classification. in this case we have chosen the Mel-Frequency Cepstral Coefficient (MFCC). The classification process is based on three machine learning supervised classifiers: -Support Vector Machine (SVM) -Knearest neighbors (KNN) - Decision tree (DT). Our proposed system evaluated using TOS. The performance used of our system are the accuracy, sensitivity, and specificity, F1 score, and Receiver Operating Characteristics (ROC).