COVID 19 Patient Recognition and Prevention Utilizing Machine Learning and CNN Model Techniques

The abrupt raise in COVID-19 infections is placing a pressure on the world's healthcare systems. At this point, it is crucial to perform a quick, precise, and untimely clinical estimation of disease severity. Thus, it has become essential but challenging to be able to spot situations that are about to die. To identify accurate and substantial predictors of mortality risk in these circumstances, we can analyze blood samples and chest X-ray pictures from patients. To recognize the majority accurate biomarkers of patient humanity, a mathematical modeling technique depending on cutting-edge interpretable machine learning methods can be developed. Five biomarkers-age, Blood Urea Nitrogen (BUN), C - reactive protein, Creatinine, and D-dimer-can calculate a patient's mortality by an accuracy of more than 90%. Particularly D-dimer levels seem to be crucial in separating the immense most of the cases that call for urgent medical concentration. In conclusion, our study suggests a straightforward decision rule for quickly identifying high-risk patients, enabling them to be given priority and possibly lowering their fatality rates. The accuracy of the suggested Random Forest Boosted by Ada boost algorithm will be higher than that of the rival models. The ensemble model and other models are outperformed by our CNN model. Our findings demonstrate that the ensemble model's accuracy has increased to 89.818%, while the CNN model's accuracy has increased to 93.2%.