Table 6.
Summary of selected COVID-19 models including underlying methodologies, predicted features, spatial resolution, scenario analysis features, and frequency of data updates.
| Model name | Institution | URL | Methodology | Predicted featuresa | Spatial resolutionb | Scenario analysis | Frequency of data updates |
|---|---|---|---|---|---|---|---|
| COVID Forecast Hub | University of Massachusetts-Amherst Reich Lab | https://covid19forecasthub.org/ | Ensemble method combining results from multiple models | C, D, H, | N, S, C | Selected individual models in the ensemble method include scenario analysis | Weekly |
| Auquan | CDC, Auquan Data Science | https://covid19-infection-model.auquan.com/ | Fitted SD model (SEIR) | C, D | G, N, S | Limited to selected model parameters (e.g., infection spread, social distancing) | Daily |
| Columbia | Columbia Mailman School of Public Health | https://cuepi.shinyapps.io/COVID-19/ | SD model (SEIR) | C, H | S, C | Limited to adjustments to the R0 values | Daily |
| Columbia-UNC | Columbia University and UNC Chapel Hill | https://github.com/COVID19BIOSTAT/covid19_prediction | Survival-convolution model | C, D | N | NA | NA |
| IHME | University of Washington—Institute for Health Metrics and Evaluation | https://covid19.healthdata.org/united-states-of-america?view=total-deaths&tab=trend | SD model (SEIR) calibrated using real-world data | C, D, H | G, N, S | Scenario analysis based on vaccination, mask use, and government-imposed mandates | Frequently |
| DDS | University of Texas at Austin UT | https://dds-covid19.github.io/index.html | Negative binomial linear dynamic system | C, D | N, S | NA | NA |
| Google-HSPH | Google Cloud AI | https://datastudio.google.com/c/reporting/52f6e744-66c6-47aa-83db-f74201a7c4df/page/EfwUB | Combination of SD model (SEIR) and covariates encoding within a computational graph framework | C, D, H | S, C | NA | Bi-weekly |
| ISU | Iowa State University | https://covid19.stat.iastate.edu/ | Discrete-time spatial epidemic model | C, D | S, C | NA | Daily |
| JHU-APL | John Hopkins University Applied Physics Laboratory LLC | https://buckymodel.com/ | Spatially distributed SD models (SEIR) stratified based on age | C, D, H | S, C | NA | NA |
| MIT-ORC | Massachusetts Institute of Technology Operations Research Center | https://www.covidanalytics.io/projections | Adjusted SD model (SEIR) | C, D, H | G, N, S | NA | NA |
| Northeastern—MOBS | Northeastern University | https://covid19.gleamproject.org/ | Adjusted SD model (SEIR) using a metapopulation approach and age-specific contact matrix | C, D, H | N, S | Scenario analysis based on different levels of social distancing | Weekly |
| Oliver Wyman | Oliver Wyman | https://pandemicnavigator.oliverwyman.com/ | Extended SD model (SIR) including detected and undetected infected populations | C, D | G, N, S, C | Scenario analysis based on mobility and testing | Daily |
| UCLA | University of California LA | https://covid19.uclaml.org/ | Adjusted SD model (SEIR) accounting for unreported recovery | C, D | G, N, S | NA | Weekly |
| UCSB | University of California Santa Barbara | https://github.com/Gandor26/covid-open/ | Attention crossing time series | C | S | NA | Weekly |
| UGA—CEID | University of Georgia Center for the Ecology of Infectious Disease | https://github.com/cdcepi/COVID-19-Forecasts/blob/master/COVID-19_Forecast_Model_Descriptions.md#Auquan | Statistical Random Walk Model | C, D | N, S, C | NA | Weekly |
| UT | University of Texas | https://covid-19.tacc.utexas.edu/projections/ | Ensemble of curve fitting and SD model (SEIR) | D | S | NA | Daily |
aC Case prediction, D death prediction, H hospitalization prediction.
bG Global-level predictions (i.e., different countries), N national-level predictions, S state-level predictions, C county-level predictions.