Machine learning for air quality forecasting: Insights from five provinces of Rwanda
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Accurately predicting air quality is a crucial challenge for public health and environmental management. This study compares and contrasts machine learning approaches to benchmark best practices for the Rwandan context and to evaluate the added value of advanced statistical methods for air quality monitoring in data-scarce settings. We forecast fine particulate matter (PM 2.5 ) concentrations across five provinces in Rwanda, using multi-year meteorological and air quality data to identify context-specific patterns. This work establishes a methodological foundation for context-optimized early warning systems and informs policy interventions to improve air quality management in Rwanda. By rigorously testing machine learning capabilities against regional constraints, we demonstrate how machine learning can reduce population exposure to pollution, quantify attribution gaps in under-monitored regions, and improve sustainable environmental governance in resource-limited settings. The results indicate significant seasonal variability, with higher PM 2.5 levels during dry seasons than wet seasons. Our evaluation demonstrates that machine learning models can capture complex, non-linear relationships between environmental variables and pollution trends, although performance varies between algorithms. Limitations remain, including the integration of real-time data streams and localized variables such as industrial emissions, road traffic, and agricultural practices.
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.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it