Prediction and Analysis of Corona Virus Disease (COVID-19) using Cubist and OneR

Abstract

Inflammatory syndrome usually occurs with the adults who have already affected with corona virus. Multi system syndrome in children may affect different parts of the body like lungs, liver, kidneys and brain. The level of current infection can be known by proceeding with two kinds of tests, namely viral tests and anti body tests. Viral tests determine the level of current infection and antibody tests are about past infection. Though decision on test will be provided by health care units, there will be an uncertainty in evaluation of results. The variation in potential of antibodies will influence the resulting parameters. The effect of inaccurate test results is to be traced, because people may suffer from reinfection in their bodies over a period of time. In this paper the uncertainty analysis is carried out using Cubist and OneR algorithms. In the cubist algorithm leaf nodes can be analyzed using regression models. Accuracy of the predictions are analyzed using contingency table and false positives and true negatives are tracked using confusion matrix. This analysis assists in determining the trans conditional state of the disease with dynamism. With the analysis carried out cubist proves to be the best algorithm.