COVID-19 X-ray Prediction Using Machine Learning Models

In this paper, we aim to predict whether the patients are COVID-inflected or not using machine learning models, including AlexNet, VGG, YOLOv5, and ResNet18. Since the X-ray images are large for processing, we utilize image decomposition and selective search to decouple the images and then send them to the network to predict the label. We analyze the available dataset, which concludes that the VGG network obtains the best performance with 86.84%. Furthermore, we find that the detection process in image pre-processing can help the classification method to handle the prediction task.