Table 1 Number of variables used from respective sources with some examples given, complete list with distributions given in supplementary material.
From: Ensemble machine learning of factors influencing COVID-19 across US counties
Source | N Var. | Var. Examples |
|---|---|---|
USAFacts | 6 | COVID-19 outcome data, population |
Bureau of Economic Analysis (BEA) | 1 | GDP |
5-Year American Community Survey (ACS), 2014–2018 | 14 | County percentages by Sex and Ethnicity, Employment, Household Income, use of Public Transportation |
TIGER/Line Geodatabases | 7 | Latitude, longtitude, land area |
TIGER/Line Geodatabases; Federal Aviation Administration (FAA) | Â | Distance to Airports |
Interactive Atlas of Heart Disease and Stroke (2014–2016) | 4 | Number of Hospitals, Stroke, Access to Parks |
County Health Rankings and Roadmaps | 21 | Life Expectancy, Smoking, Obesity,, Food Access, Mental Health, Physicians, Houshold Overcrowding etc. |
Centers for Medicare & Medicaid Services (CMS) | 15 | Druge Abuse, Hypertension, Hyperlipidemia, Osteoporosis, etc. |
National Centers for Environmental Information | 1 | Precipitation |
CDC’s Social Vulnerability Index (SVI) | 11 | Percentile over 65 or under 17, Minority Scores, Limited English, Low Income Housing Estimates, Number Institutionalized |
Quarterly Census of Employment and Wages | 14 | Labor force types, farming/mining, private industry, education/healthcare etc. |
MIT election lab | 1 | Calculated Proportion Voted Republican 2016 |
6 | Google mobility to location type, Residence, Grocery etc. |
