Role of Classification Model with Fuzzy Model to Predict Covid-19: A Comparative Study

Arya, L. V.; Singh, P.

The ongoing Coronavirus pandemic is the infectious disease brought about by the latest discovered coronavirus, and is affecting many countries globally. At the present time it is difficult to test everybody globally so a model is developed in this paper that can help in predicting the risk of coronavirus. A decision tree is constructed for this purpose based on certain attributes like fever, dry cough, tiredness, difficulty to breath, and chest pain;and is compared using various other methods. Classification Functions are used for the prediction and the results are compared based on accuracy. It was observed that Multilayer Perceptron classifier achieved the highest accuracy of 95.31%, however the generated tree using J48 achieved same accuracy (90.63%) as that of LMT & Logistic classifiers. Rules generated by Decision tree are then used in the Fuzzy Inference System using MATLAB to give predictions related to the disease risk. © 2021, Springer Nature Switzerland AG.