id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
cord-277862-yl7m77fo	Li, M.	Identifying novel factors associated with COVID-19 transmission and fatality using the machine learning approach	2020-06-12		.txt	text/plain	3935	228	54	To uncover the critical factors for COVID-19 risk within a single country, we used logistic models to predict COVID-19 cases, deaths, and CFRs in the 50 U.S. states. Consistent with the previous results, medium temperature, arid climate, social distancing (major sports events), per capita income, longitude, and the average age of childbirth were positive predictors of COVID-19 cases, deaths, and/or CFRs, and humidity, smoking rate, and international tourism revenue were negative predictors. To compare the relative contribution of high and low temperature in predicting COVID-19 risk, we built logistic models with both variables to predict COVID-19 cases, deaths, and CFRs in the 154 countries. Second, we identified novel factors associated with COVID19, including the unitary state governing system as a positive predictor of COVID-19 cases and deaths, blood type B as a protective factor for All rights reserved.	./cache/cord-277862-yl7m77fo.txt	./txt/cord-277862-yl7m77fo.txt