Improving COVID-19 Detection: Comparative Performance Analysis of Machine Learning and Deep Learning Algorithms using CT Scan Images

Abstract

For the disease to be managed and controlled effectively, COVID-19 instances must be accurately and quickly identified. In this investigation, we look at the application of CT scan pictures for COVID-19 illness identification and assess the effectiveness of several deep learning and machine learning techniques. KNN, SVM, Decision Tree, Nave Bayes, Logistic Regression, CNN, LSTM, InceptionV3, ResNet50, and MobileNet were trained and assessed using Using CT scan image collection and preprocessing, measures such as precision, accuracy, recall, F-score, and false prediction rate are calculated. Our findings demonstrate the extraordinary performance of the CNN algorithm, which achieved 100% precision, 100% F-score, 100% recall, a100% accuracy, and a 0% false prediction rate. Both the KNN and the SVM algorithms demonstrated promising outcomes, according to the comparative analysis. These results demonstrate that machine learning and deep learning methods have the capacity to correctly identify COVID-19 instances from CT scan pictures. A future study may investigate ensemble models, transfer learning strategies, and the integration of several modalities to enhance the diagnostic precision and generalizability of the algorithms.