Machine Learning Applications in Population and Public Health: Guidelines for Development, Testing, and Implementation
Bibliographic record
Abstract
Unlabelled: Machine learning (ML), a subset of artificial intelligence, uses large datasets to identify patterns between potential predictors and outcomes. ML involves iterative learning from data and is increasingly used in population and public health. Examples include early warning of infectious disease outbreaks, predicting the future burden of noncommunicable diseases, and assessing public health interventions. However, ML can inadvertently produce biased outputs related to the quality and quantity of data, who is engaged and helping direct the analysis, and how findings are interpreted. Specific guidelines for using ML in population and public health have not yet been created. We assembled a diverse team of experts in computer science, statistical modeling, clinical and population health epidemiology, health economics, ethics, sociology, and public health. Drawing on literature reviews and a modified Delphi process, we identified five key recommendations: (1) prioritize partnerships and interventions to support communities considered structurally disadvantaged; (2) use ML for dynamic situations, such as public health emergencies, while adhering to ethical standards; (3) conduct risk assessments and bias mitigation strategies aligned with identified risks; (4) ensure technical transparency and reproducibility by publicly sharing data sources and methodologies; and (5) foster multidisciplinary dialogue to discuss the potential harms of ML-related bias and raise awareness among the public and public health community. The proposed guidelines provide operational steps for stakeholders, ensuring that ML tools are not only effective but also ethically grounded and feasible in real-world scenarios.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.215 | 0.400 |
| Meta-epidemiology (narrow) | 0.004 | 0.003 |
| Meta-epidemiology (broad) | 0.003 | 0.005 |
| Bibliometrics | 0.011 | 0.009 |
| Science and technology studies | 0.003 | 0.009 |
| Scholarly communication | 0.010 | 0.010 |
| Open science | 0.013 | 0.011 |
| Research integrity | 0.018 | 0.018 |
| Insufficient payload (model declined to judge) | 0.017 | 0.023 |
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; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
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".