ABSTRACT
COVID-19 prediction models were developed by using the patient's symptoms and medical history as input data and comparison of classification models is done. Decision Tree (DT) and Random Forest (RF) models are the most accurate machine learning classification models among all the classifiers used for this type of dataset giving the accuracy of 98.5 percent. Gaussian Naïve Bayes and Logistic Regression are not very accurate models for predicting COVID-19 cases based on symptoms and medical histories. Breathing Problem, Sore Throat, Dry cough and fever are the most important symptoms to predict whether the patient is having the COVID-19 infection or not. Also, in all these symptoms Breathing Problem has the highest weightage.
