An Efficient COVID-19 Mortality Risk Prediction Model Using Deep Synthetic Minority Oversampling Technique and Convolution Neural Networks

The COVID-19 virus has made a huge impact on people’s lives ever since the outbreak happened in December 2019. Unfortunately, not completely vanished from world yet, and thus, global agitation is still increasing with mutations variants of same. Early diagnosis best way to decline mortality risk associated it. This urges necessity developing new computational approaches that can analyze large dataset predict disease time. Currently, automated major area research for accurate timely predictions. Artificial intelligent (AI)-based techniques such as machine learning (ML) deep (DL) be deployed this purpose. In this, compared traditional techniques, Learning show prominent results. Yet it requires optimization terms complex space problems. To address issue, proposed method combines predictive models convolutional neural network (CNN), long short-term memory (LSTM), auto-encoder (AE), cross-validation (CV), synthetic minority oversampling (SMOTE). proposes six different combinations forecasting CV-CNN, CV-LSTM+CNN, IMG-CNN, AE+CV-CNN, SMOTE-CV-LSTM, SMOTE-CV-CNN. performance each model evaluated using various metrics standard approved by Montefiore Medical Center/Albert Einstein College Medicine Institutional Review Board. experimental results SMOTE-CV-CNN outperforms other achieving an accuracy 98.29%. Moreover, been existing prediction methods based both (DL), demonstrated superior accuracy. Based analysis, inferred ability effectively related COVID-19.