Prediction of SARS-COVID-19 Based on Transfer Machine Learning Techniques Using Lungs CT Scan Images

SARS-COVID-19 is a worldwide pandemic, which affected almost every country in the world. It spread so quickly as the results caused millions of deaths all over the world. Machine learning and deep learning techniques are playing an important role in medical diagnosis. Several machine learning and deep learning techniques are proposed for the prediction of SARS-COVID-19. The analysis of SARS-COVID is performed based on the lung images. These lung images can be got via CT scans of the lungs or X-ray images of the lungs. As the data formed via CT scan of lungs present in the form of images, machine learning and deep learning techniques are taking the help of convolutional neural networks. In this research paper, efficient machine learning and deep learning-based, transfer learning approach for the prediction of SARS-COVID-19 is proposed. The datasets are collected from different sources which are based on computer tomography (CT) scan images of the lungs, as they provide more accurate results. A total of 3227 images of both COVID positive and COVID negative patients were taken for the experimentation. The proposed model achieved a training accuracy of 98% and a validation accuracy of 96% is provided.