Cleaning up the Big Smoke: Forecasting London’s Air Pollution Levels Using Energy-Efficient AI
Bibliographic record
Abstract
Air pollution exposure poses a major risk to human health, with devastating effects ranging from causing respiratory and cardiovascular diseases, to adverse impacts on cognitive abilities, mental health, and prenatal development.In the case of an excessive build-up of air contaminants, emergency measures must be enacted to reduce human exposure and decrease pollution levels.Hence, cities worldwide have invested in sophisticated air pollution monitoring systems to assess pollution levels and inform public health advice.Predicting spikes in air pollution a few hours in advance is critical in reducing human exposure as much as possible.While deep neural networks have become popular for this task, standard machine learning approaches remain very attractive: they deliver competitive performance without relying on specialised equipment and consume much less energy than their deep learning counterparts.Experiments conducted on London air quality data demonstrate that Linear Regression achieves stateof-the-art performance, with 1-hour and 24-hour predictions displaying 0.2 and 3.2 mean absolute errors respectively.Moreover, its energy usage is a fraction of that of its deep learning competitor, LSTM, consuming over 2000 times less energy for training, and over 100 times less energy for prediction.The results demonstrate that standard machine learning approaches can provide an accurate and energyefficient approach to air pollution forecasting, without prohibitive hardware investments.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 source (direct Gemma or distilled Codex), 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".