Application of Machine Learning to Two Large-Sample Household Travel Surveys: A Characterization of Travel Modes
Bibliographic record
Abstract
Even in a context of rapidly evolving transportation and information technologies, household travel surveys remain an essential source of information for transportation planning. Moreover, as planning authorities become increasingly concerned with reducing the use of the private car, travelers’ mode choice patterns should be reexamined. In this study, a machine learning algorithm (Random Forest) was employed to characterize the use of eight different travel modes observed in two consecutive household travel surveys undertaken in Montreal, Canada. The analysis incorporated roughly 160,000 observed trips. The Random Forest algorithm was trained on the 2008 survey data and applied to the 2013 survey. The usefulness of the algorithm was evaluated using two numerical representations: the confusion matrix and the importance matrix. The results of this evaluation showed that the Random Forest algorithm could generate a detailed and precise characterization of travel submarkets for four of the most commonly observed modes of travel (auto-drive, public transit, school bus, and walk) using 11 attributes of households, persons, and trips. However, the auto-passenger mode was difficult to characterize because of its dependence on unobserved intra-household interactions. The algorithm also had difficulty identifying users of rarely observed modes (park-and-ride, kiss-and-ride, bicycle), but performed better in this regard than a traditional mode choice model. Finally, traveler’s age and the spatial orientation of origin–destination pairs were found to be decisive factors in the use of the auto-drive mode. This finding, combined with the stability of mode choice patterns observed over 5 years, highlights the difficulty of significantly reducing automobile use.
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 | 0.002 | 0.010 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 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 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".