Performance Evaluation of Multi-class Classification based Detection of CoViD-19 using Machine Learning Algorithms

Abstract

The Corona Virus Disease (CoViD-19) has underscored the need for accurate and rapid diagnostic tools to combat the spread of the virus. Machine learning techniques have shown promise in detecting CoViD-19 from medical imaging data, such as X-rays and CT scans. This study presents a comprehensive evaluation of multi-class classification methods for the detection of CoViD-19 using adverse dataset of medical images. Our research focuses on the development and assessment of machine learning algorithms capable of classifying CoViD-19 cases into multiple categories, including distinguishing CoViD-19 from other respiratory conditions and providing insights into the disease's severity. We utilize a dataset comprising a wide range of chest X-ray and CT scan images obtained from various sources and demographic groups to ensure the robustness of our model. In this study, we explore and compare the performance of several machine learning algorithms, including Decision Tree, Support Vector Machines (SVMs) and Naïve Bayes among others. We also employ data preprocessing techniques, feature selection, and data augmentation to enhance the model's accuracy and generalization capabilities. Furthermore, we present an in-depth analysis of the model's performance metrics, such as Accuracy, Specificity, Precision, Recall, Time, and Error rate to provide a comprehensive evaluation of its diagnostic capabilities. Our results show case the strengths and weaknesses of each algorithm in the context of CoViD-19 detection and multi-class classification. The findings of this research are valuable for healthcare practitioners, researchers, and policy makers involved in the fight against CoViD-19. By identifying the most effective machine learning algorithms and techniques for multi-class classification-based CoViD-19 detection, we aim to contribute to the development of robust and reliable diagnostic tools, ultimately aiding in the early identification and management of CoViD-19cases.