Mortality prediction of COVID-19 patients using soft voting classifier

• Published in: International journal of cognitive computing in engineering Vol. 3; pp. 172 - 179
• Main Authors: Rai, Nishant, Kaushik, Naman, Kumar, Deepika, Raj, Chandan, Ali, Ahad
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.06.2022; The Authors. Publishing Services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd; KeAi Communications Co., Ltd
• Subjects: Classification; COVID19; Ensemble; Mortality; Soft voting; COVID19; Soft voting; Mortality; Ensemble; Classification
• ISSN: 2666-3074; 2666-3074
• DOI: 10.1016/j.ijcce.2022.09.001

•The COVID outbreak of 2019 is the biggest crisis human civilisation has faced in recent history. A virus called Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) was the cause of this outbreak.•Various techniques have been used in multiple studies to predict the severity and mortality of patients ranging from their race, chest CT scans, medical biomarkers, and past medical records.•The model proposed in the study determines the probability of death of a patient admitted to a hospital taking into account various aspects of the patient such as age and the blood biomarkers.•The proposed model uses a soft voting classifier to predict the mortality of the patient admitted to the hospital with base estimators such as Random Forest, XGBoost, Gradient Boosting Classifier, and Extra Tree Classifier.•Accuracy, Precision, Recall, and F1 Score have been used as the evaluation criteria for checking the robustness of the proposed methodology.•Comparing the proposed model with base models and the existing state of the art models provides a superior F1 score which is crucial during the COVID waves. COVID-19 is a novel coronavirus that spread around the globe with the initial reports coming from Wuhan, China, turned into a pandemic and caused enormous casualties. Various countries have faced multiple COVID spikes which put the medical infrastructure of these countries under immense pressure with third-world countries being hit the hardest. It can be thus concluded that determining the likeliness of death of a patient helps in avoiding fatalities which inspired the authors to research the topic. There are various ways to approach the problem such as past medical records, chest X-rays, CT scans, and blood biomarkers. Since blood biomarkers are most easily available in emergency scenarios, blood biomarkers were used as the features for the model. The data was first imputed and the training data was oversampled to avoid class imbalance in the model training. The model is composed of a voting classifier that takes in outputs from multiple classifiers. The model was then compared to base models such as Random Forest, XGBoost, and Extra Trees Classifier on multiple evaluation criteria. The F1 score was the concerned evaluation criterion as it maximizes the use of the medical infrastructure with the minimum possible casualties by maximizing true positives and minimizing false negatives.