Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods

Abstract Background Coronavirus disease (COVID-19) presents an unprecedented threat to global health worldwide. Accurately predicting the mortality risk among the infected individuals is crucial for prioritizing medical care and mitigating the healthcare system’s burden. The present study aimed to a...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Cindy Feng, George Kephart, Elizabeth Juarez-Colunga
Formato:	article
Lenguaje:	EN
Publicado:	BMC 2021
Materias:	COVID-19 mortality Predictive model Generalized additive model Classification trees Extreme gradient boosting Medicine (General) R5-920
Acceso en línea:	https://doaj.org/article/e95c431b7b8c4736b3daa26d7bd17fa1
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:e95c431b7b8c4736b3daa26d7bd17fa1
record_format	dspace
spelling	oai:doaj.org-article:e95c431b7b8c4736b3daa26d7bd17fa12021-11-28T12:38:47ZPredicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods10.1186/s12874-021-01441-41471-2288https://doaj.org/article/e95c431b7b8c4736b3daa26d7bd17fa12021-11-01T00:00:00Zhttps://doi.org/10.1186/s12874-021-01441-4https://doaj.org/toc/1471-2288Abstract Background Coronavirus disease (COVID-19) presents an unprecedented threat to global health worldwide. Accurately predicting the mortality risk among the infected individuals is crucial for prioritizing medical care and mitigating the healthcare system’s burden. The present study aimed to assess the predictive accuracy of machine learning methods to predict the COVID-19 mortality risk. Methods We compared the performance of classification tree, random forest (RF), extreme gradient boosting (XGBoost), logistic regression, generalized additive model (GAM) and linear discriminant analysis (LDA) to predict the mortality risk among 49,216 COVID-19 positive cases in Toronto, Canada, reported from March 1 to December 10, 2020. We used repeated split-sample validation and k-steps-ahead forecasting validation. Predictive models were estimated using training samples, and predictive accuracy of the methods for the testing samples was assessed using the area under the receiver operating characteristic curve, Brier’s score, calibration intercept and calibration slope. Results We found XGBoost is highly discriminative, with an AUC of 0.9669 and has superior performance over conventional tree-based methods, i.e., classification tree or RF methods for predicting COVID-19 mortality risk. Regression-based methods (logistic, GAM and LASSO) had comparable performance to the XGBoost with slightly lower AUCs and higher Brier’s scores. Conclusions XGBoost offers superior performance over conventional tree-based methods and minor improvement over regression-based methods for predicting COVID-19 mortality risk in the study population.Cindy FengGeorge KephartElizabeth Juarez-ColungaBMCarticleCOVID-19 mortalityPredictive modelGeneralized additive modelClassification treesExtreme gradient boostingMedicine (General)R5-920ENBMC Medical Research Methodology, Vol 21, Iss 1, Pp 1-14 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	COVID-19 mortality Predictive model Generalized additive model Classification trees Extreme gradient boosting Medicine (General) R5-920
spellingShingle	COVID-19 mortality Predictive model Generalized additive model Classification trees Extreme gradient boosting Medicine (General) R5-920 Cindy Feng George Kephart Elizabeth Juarez-Colunga Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods
description	Abstract Background Coronavirus disease (COVID-19) presents an unprecedented threat to global health worldwide. Accurately predicting the mortality risk among the infected individuals is crucial for prioritizing medical care and mitigating the healthcare system’s burden. The present study aimed to assess the predictive accuracy of machine learning methods to predict the COVID-19 mortality risk. Methods We compared the performance of classification tree, random forest (RF), extreme gradient boosting (XGBoost), logistic regression, generalized additive model (GAM) and linear discriminant analysis (LDA) to predict the mortality risk among 49,216 COVID-19 positive cases in Toronto, Canada, reported from March 1 to December 10, 2020. We used repeated split-sample validation and k-steps-ahead forecasting validation. Predictive models were estimated using training samples, and predictive accuracy of the methods for the testing samples was assessed using the area under the receiver operating characteristic curve, Brier’s score, calibration intercept and calibration slope. Results We found XGBoost is highly discriminative, with an AUC of 0.9669 and has superior performance over conventional tree-based methods, i.e., classification tree or RF methods for predicting COVID-19 mortality risk. Regression-based methods (logistic, GAM and LASSO) had comparable performance to the XGBoost with slightly lower AUCs and higher Brier’s scores. Conclusions XGBoost offers superior performance over conventional tree-based methods and minor improvement over regression-based methods for predicting COVID-19 mortality risk in the study population.
format	article
author	Cindy Feng George Kephart Elizabeth Juarez-Colunga
author_facet	Cindy Feng George Kephart Elizabeth Juarez-Colunga
author_sort	Cindy Feng
title	Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods
title_short	Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods
title_full	Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods
title_fullStr	Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods
title_full_unstemmed	Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods
title_sort	predicting covid-19 mortality risk in toronto, canada: a comparison of tree-based and regression-based machine learning methods
publisher	BMC
publishDate	2021
url	https://doaj.org/article/e95c431b7b8c4736b3daa26d7bd17fa1
work_keys_str_mv	AT cindyfeng predictingcovid19mortalityriskintorontocanadaacomparisonoftreebasedandregressionbasedmachinelearningmethods AT georgekephart predictingcovid19mortalityriskintorontocanadaacomparisonoftreebasedandregressionbasedmachinelearningmethods AT elizabethjuarezcolunga predictingcovid19mortalityriskintorontocanadaacomparisonoftreebasedandregressionbasedmachinelearningmethods
_version_	1718407865551552512

Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methods

Ejemplares similares