Performance Analysis of Machine Learning Algorithms for COVID-19 Detection

EasyChair Preprint 9496

Abstract

The Novel Coronavirus Illness 2019 (COVID-19) is a deadly infectious disease that was discovered for the first time in December 2019 in Wuhan, Hubei, China, and has since spread globally. The patient's life is at risk when the patient's corona illness became severe. The patient's lungs are where the coronavirus typically attacks. These days, diagnostic kits only look for viral infections, which deceives doctors while treating patients. The essential organs of patients with less infection suffer needless harm when all patients get the same treatment. The proper non-invasive treatment for infected patients is provided in this publication. An important tool for both investigation and prediction of COVID-19 patients is the examination of the coronavirus by dissecting chest X-ray images. We provide a hybrid approach-based automated method for detecting Covid. We are detecting Covid using CNN and SVM algorithms. The X-ray images are not consistent; hence the feature extraction process uses the CNN method. Before CNN, we employed a technique called data augmentation to create a suitable training dataset. Data augmentation aids in increasing the quantity of the data set for effective training as well as the quality of the image dataset. Because SVM is tolerant of feature variances, it is utilized for classification. The primary objective is to provide a framework that aids clinical doctors in determining the severity of a chest infection so that appropriate, life-saving therapy can be given. The application of deep learning and machine learning-based tools will detect the degree of infection in the chest and lead to proper therapy and avoid expensive treatment for all patients as the current diagnosis procedure is time-consuming.

Keyphrases: Artificial Intelligence, COVID-19, Convolutional Neural Network, SVM, X-rays

@booklet{EasyChair:9496,
  author    = {Anuradha D. Thakare and Pranjali G. Gulhane and Shruti M. Chaudhari and Hemant Baradkar},
  title     = {Performance Analysis of Machine Learning  Algorithms for COVID-19 Detection},
  howpublished = {EasyChair Preprint 9496},
  year      = {EasyChair, 2022}}