Dataset Information

A seq2seq model to forecast the COVID-19 cases, deaths and reproductive R numbers in US counties.

ABSTRACT: The global pandemic of coronavirus disease 2019 (COVID-19) has killed almost two million people worldwide and over 400 thousand in the United States (US). As the pandemic evolves, informed policy-making and strategic resource allocation relies on accurate forecasts. To predict the spread of the virus within US counties, we curated an array of county-level demographic and COVID-19-relevant health risk factors. In combination with the county-level case and death numbers curated by John Hopkins university, we developed a forecasting model using deep learning (DL). We implemented an autoencoder-based Seq2Seq model with gated recurrent units (GRUs) in the deep recurrent layers. We trained the model to predict future incident cases, deaths and the reproductive number, R . For most counties, it makes accurate predictions of new incident cases, deaths and R values, up to 30 days in the future. Our framework can also be used to predict other targets that are useful indices for policymaking, for example hospitalization or the occupancy of intensive care units. Our DL framework is publicly available on GitHub and can be adapted for other indices of the COVID-19 spread. We hope that our forecasts and model can help local governments in the continued fight against COVID-19.

SUBMITTER: Zhang-James Y

PROVIDER: S-EPMC8132245 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:On March 13, 2020, the World Health Organization (WHO) declared the 2019 coronavirus disease (COVID-19) caused by the novel coronavirus SARS-CoV2 a pandemic. Since then the virus has infected over 9.1 million individuals and resulted in over 470,000 deaths worldwide (as of June 24, 2020). Here, we discuss the spatial correlation between county population health rankings and the incidence of COVID-19 cases and COVID-19 related deaths in the United States. We analyzed the spread of the disease based on multiple variables at the county level, using publicly available data on the numbers of confirmed cases and deaths, intensive care unit beds and socio-demographic, and healthcare resources in the U.S. Our results indicate substantial geographical variations in the distribution of COVID-19 cases and deaths across the US counties. There was significant positive global spatial correlation between the percentage of Black Americans and cases of COVID-19 (Moran I = 0.174 and 0.264, p < 0.0001). A similar result was found for the global spatial correlation between the percentage of Black American and deaths due to COVID-19 at the county level in the U.S. (Moran I = 0.264, p < 0.0001). There was no significant spatial correlation between the Hispanic population and COVID-19 cases and deaths; however, a higher percentage of non-Hispanic white was significantly negatively spatially correlated with cases (Moran I = -0.203, p < 0.0001) and deaths (Moran I = -0.137, p < 0.0001) from the disease. This study showed significant but weak spatial autocorrelation between the number of intensive care unit beds and COVID-19 cases (Moran I = 0.08, p < 0.0001) and deaths (Moran I = 0.15, p < 0.0001), respectively. These findings provide more detail into the interplay between the infectious disease and healthcare-related characteristics of the population. Only by understanding these relationships will it be possible to mitigate the rate of spread and severity of the disease.

Dataset Information

A seq2seq model to forecast the COVID-19 cases, deaths and reproductive R numbers in US counties.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets