A Performance Comparison Between Machine Learning, Deep Learning, and Bayesian Approaches for COVID-19 Diagnosis

The COVID-19 pandemic emphasizes the critical need for rapid and reliable diagnostic procedures to improve infection detection and prevention. This study compares Machine Learning (ML), Deep Learning(DL) and probabilistic graphical models (PGMs) for predicting COVID-19 infection using a publicly available dataset. ML models such as Support Vector Machine(SVM), Random Forest(RF), Decision Tree(DT), K-Nearest Neighbour(KNN), Multi Layer Perceptron(MLP) and Voting Classifier(VC) with Synthetic Minority Over sampling Technique (SMOTE) attained an accuracy of 98.9%. The Deep Neural Network (DNN) attained 97.88% accuracy and Bayesian Network(BN)implemented using Probabilistic Graphical Models in Python(PGMPY) achieved 96 % accuracy. Despite lower accuracy, BN helps in complex diagnostic tasks. It also demonstrates the power of ensemble ML and DL models. Therefore these results reflect the advantages attained from ensemble ML,DL and BN models in terms of flexibility and the ability of each approach to solve complex diagnostic problems and tasks, as represented by the predictive modeling of COVID-19 cases.