Identifying Significant Predictors of COVID-19 Mortality Rate in the United States

Sarah Whelchel, Western Washinton University

Description

Identifying useful predictors of COVID-19 mortality rate is of critical importance to fight the ongoing pandemic. Using the USA daily state report for April 14, 2020 in the JHU CSSE COVID-19 Dataset, we assess the significance of predictors possibly associated with mortality rate, including latitude, longitude, incident rate, testing rate, and hospitalization rate. Our statistical analysis using the Lasso regression and linear discriminant analysis (LDA) suggest that while hospitalization rate seems to be well correlated with mortality rate, latitude, longitude, incident rate, and testing rate do not apear to be useful as predictors. We further discuss possible causes and implications of the results above by analyzing associations between testing rate, incident rate, and mortality rate.

 
May 18th, 9:00 AM May 22nd, 5:00 PM

Identifying Significant Predictors of COVID-19 Mortality Rate in the United States

Identifying useful predictors of COVID-19 mortality rate is of critical importance to fight the ongoing pandemic. Using the USA daily state report for April 14, 2020 in the JHU CSSE COVID-19 Dataset, we assess the significance of predictors possibly associated with mortality rate, including latitude, longitude, incident rate, testing rate, and hospitalization rate. Our statistical analysis using the Lasso regression and linear discriminant analysis (LDA) suggest that while hospitalization rate seems to be well correlated with mortality rate, latitude, longitude, incident rate, and testing rate do not apear to be useful as predictors. We further discuss possible causes and implications of the results above by analyzing associations between testing rate, incident rate, and mortality rate.