OU researchers have created new data models that can better predict the spread of COVID-19 in rural areas.
According to a university press release, researchers, led by Gallogly College of Engineering Associate Professor Charles Nicholson, have applied machine learning techniques and data — such as demographics, social media, health factors and mobility — to create new research models that will provide better insight on the spread of COVID-19. The new models of research will be community-specific and designed for smaller populations.
“We wanted to primarily focus on developing models for rural areas and counties for whom the existing forecasting models were limited or inaccurate due to small sample sizes,” Nicholson said in the release. “Our ultimate goal is to provide early warning and actionable information to medical service providers.”
Due to the rapid spread of COVID-19, it has been difficult to create accurate models, particularly in the earlier months of the outbreak, according to the release. Nicholson claimed most of the data from the first few months skewed toward more populated and urban areas. According to Nicholson, these more localized models will help policymakers make better-informed decisions in terms of COVID-19 prevention in their communities.
Additionally, county-level research models will help smaller communities understand how many resources they need to ensure that they are not overextending their finances, according to the release. As larger cities have more resources readily available, it is harder for rural areas to project their needs.
“For example, Oklahoma’s budget per capita is only 52.6 percent of New York’s. Accurate disease projections are potentially ‘mission critical’ for Oklahoma given this relative shortfall of resources,” Nicholson said in the release.