الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 88 |  |  |
| | | from | 88 | | |
Abstract
The COVID-19 pandemic has underscored the importance of rapid, accurate, and scalable methods for disease monitoring, forecasting, and management. Machine learning techniques have emerged as invaluable tools in addressing various facets of the pandemic, ranging from diagnosing the virus, predicting its spread, optimizing resource allocation, to drug discovery and vaccine development. This thesis explores the intersection of COVID-19 and machine learning, highlighting pivotal advancements, challenges, and future directions in leveraging data-driven approaches to combat and mitigate the impacts of infectious diseases like COVID-19. This thesis proposes a novel COVID-19 detection methodology using machine learning techniques. All feature extractors have been measured and tested. CADS is divided into two stages: (I) the features extraction and selection stage and (ii) the disease detection stage. To increase classification accuracy, two statistical first-order features were extracted and added to the patient’s blood test data in the first step, and then the grey wolf Optimizer (GWO) was used to choose the most significant features. The selected features are next categorized as normal or COVID-19 using a machine learning classification technique; the performance of the K-Nearest Neighbor (KNN) and support vector machine (SVM) algorithms were compared in this phase. As a result, KNN and SVM are the most commonly used classifiers in the medical industry, with SVM having the greatest accuracy rate of more than %95.