Research Article Open Access

Death Rate Estimation Model for COVID-19 Patients Using CIXCNet Classifier

Abstract

The abrupt rise of Coronavirus Disease (COVID-19) infected cases causes huge pressure on healthcare sectors throughout the world. In this scenario, an earlier and accurate detection of the disease is essential. It is necessary to identify robust and meaningful markers of mortality risk of COVID-19 patients. An interpretable Machine Learning (ML) algorithm has to be designed to detect the most distinguishing biomarkers of patient death. It should aim to distinguish the patients at imminent risk, thereby relieving clinical burden and potentially reducing the mortality rate. This paper aims to design a death rate estimation model for COVID-19 patients to detectthe vital biomarkers causing death. In this model, CIXCNetclassifier is developed in whichConvolution Neural Network (CNN) and XGBoost with Cohart Intelligence (CI) optimization are combined. In traditional classifiers, the structure of the model is determined by hyperparameters, which is time-consuming due to manual tuning of the parameters. Hence the CI optimization method is employed for tuning the hyper parameters. The model output corresponds to the patient mortality rate. Performance results show that the proposed CIXCNet classifier achieves higher accuracy, precision, recall and F1-score when compared with other classifiers.

K.Ravishankar*, Jothi Kumar C

To read the full article Download Full Article