Leveraging machine learning methods to predict COVID-19 vulnerability in U.S. counties based on socioeconomic factors
Bibliographic record
Abstract
As COVID-19 gained pandemic status, the number of confirmed cases in the US surpassed that of all other countries. Although the virus spread throughout the US, not all areas were affected equally. This retrospective study aims to explore these inequalities through pre-pandemic socioeconomic characteristics by attempting to create a predictive model for COVID-19 vulnerability at the county level. A total of 103 features of socioeconomic data for 2610 US counties (out of a total of 3007) were sourced from various online databases such as the US Census Bureau, the US Department of Agriculture, and the Association of American Medical Colleges. Additionally, to quantify each county’s COVID-19 vulnerability, we defined 3 custom measures: incidence, mortality, and case fatality. These measurements were calculated using case and death data taken 29 days after each county’s first case. Machine learning classification algorithms – including random forest, multi-layer perceptron neural network and XGBoost – were then used to predict the incidence, mortality, and case fatality of US counties. Through analysis, we were able to predict a county’s COVID-19 incidence with ~47% accuracy, mortality with ~59% accuracy, and case fatality with ~61% accuracy by looking solely at pre-pandemic socioeconomic factors. A list of important features was extracted using a built-in XGBoost function for each vulnerability measure (incidence, mortality, and case fatality). Many of these features are typically associated with pandemic spread (e.g., population density and medical infrastructure), while other features were unexpected (e.g., education) and warrant further studies to identify their role in disease propagation. Furthermore, the difficulties our model experienced support the notion that region-specific policies play an important role in successfully mitigating this crisis. The moderate success achieved in this study proves the feasibility of using classifiers as a pandemic preparedness evaluation tool.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.017 | 0.011 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.002 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".