Using machine learning to predict child active transportation prevalence
Bibliographic record
Abstract
Active school transportation (AST) can have a host of physical and mental health benefits. Unfortunately, child AST rates have declined over the last few decades. Changes to the built environment can improve AST prevalence. Due to the complexity within the road system, machine learning models may hold promise to accurately predict factors related to child AST. As such, our aim was to train and evaluate a machine learning algorithm to predict the prevalence of child AST. Data were collected from The CHASE (CHild Active-transportation Safety and the Environment) study's geodatabase, including seven Canadian municipalities/regions. The proportion of enrolled students using AST at each school was assessed by observing students arrive to school in May/June of 2018 or 2019. Data were aggregated at the school catchment zone as the unit of analysis. Both national and city-specific models were trained and validated. Root mean squared error was used to assess prediction accuracy. A measure of variable importance was also calculated. A total of 541 elementary schools were included. Median city AST prevalence ranged from 0.4 (Calgary) to 0.73 (Montreal). National and city-specific models resulted in similar prediction accuracy. Population density, Walk Score®, proportion of child population enrolled in the school, and size of residential area within each school catchment zone were frequently highly ranked in importance. Population and housing density were the two most important predictors of AST prevalence. Policies that can increase population and housing density will, therefore, likely increase AST among school-aged children. • Machine learning decision trees predicted child active transportation prevalence. • Predictors included built environment, population, and housing variables. • Models were able to predict active transportation with low error. • Housing and population density found to be highly important predictors.
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 | 0.003 | 0.009 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".