|
|
|
AUTHOR(S): Gil-Vera, Victor Daniel
|
|
TITLE Using Machine Learning to Predict the Death Risk in COVID-19 Patients |
 PDF |
|
ABSTRACT The COVID-19 pandemic caused a worldwide health crisis resulting in millions of deaths and infections, which led to the collapse of the intensive care units of many clinics and hospitals despite the strategies implemented by governments to prevent its proliferation, such as strict quarantines, social distancing, teleworking, among others. Predictive models are very useful to identify the mortality of infected patients. The objective of this study was to analyze several models used to categorize the patient's risk of passing away. According to the study's findings, the accuracy of the various models—logistic regression, K-nearest neighbors, support vector machines, Naive Bayes, decision trees, and random forest—was high (> 0.70), with random forests taking the lead (Accuracy=0.92), indicating that the models are reliable for predicting the risk of death in COVID-19 infection patients. |
|
KEYWORDS COVID-19, Databases, Limitations, Machine Learning, Predictive Models |
|
|
|
Cite this paper Gil-Vera, Victor Daniel. (2022) Using Machine Learning to Predict the Death Risk in COVID-19 Patients. International Journal of Biology and Biomedicine, 7, 56-63 |
|
|
