Data Classification Model for COVID-19 Pandemic

Folorunso, S. O.; Awotunde, J. B.; Adeboye, N. O.; Matiluko, O. E.

A significant worldwide pandemic disease that has shut the whole world’s economy and put the health care services personnel into anxiety is COronaVIrus Disease 2019 (COVID-19). It is difficult to model as it shared closely related characteristics/symptoms with other pneumonia diseases like SARS, MERS, ARDS, and Pulmonary Tuberculosis (PTB). Health practitioners use images (CT scan, Chest X-Ray (CXR)), timely occurrences (daily), audio (Cough), text (clinical and laboratory data) to detect, predict and treat patients with this disease. But machine learning has been proven by researchers when it can effectively and precisely detect, predict, classify, recommend treatment. This chapter discusses and implements a data classification task for early diagnosis and prognosis of the COVID-19 pandemic using CXR image. Classification is a supervised learning task that uses labeled data to assign items to different classes. The indicators that define a good classification task and assess classification models’ performance are Receiver Operating Characteristic (ROC), Precision-Recall Curve (PRC), Recall, F1-Score Precision. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.