Feature Selection based Artificial Intelligence Techniques for the Prediction of COVID like Diseases

Abstract

Identification of disease from therapeutic statistical evidences area single confronted task which can make a point of importance in the field of medical science. But according to the literature survey, it has been seen that still there are some chances that this challenging task can be fulfilled. In this research a feature ranking algorithm Random Forest is used for ranked the features of the attributes & later on four machine learning algorithm has been used i.e. Random forest, decision Tree, support Vector Machine & XG Boost classification algorithm to classify similar disease datasets like Jaundice, Malaria, Covid, Common cold, Typhoid, Dengue & Pneumonia. Comparison between the classifier is done on the basis of with ranking with feature selection & ranking without feature selection with the help of parameters of confusion matrix, Matthews’s correlation coefficient (MCC), area under the curve (AUC), Receiver Operating Characteristics Curve (ROC) & computational time. The results of the simulations shows the effectiveness of Covid like disease prediction is done by the feature selection ranking &classification algorithm.