Label-Less Learning for Urban Railway Transit Driver Fatigue Detection with Heart Rate Variability
Bibliographic record
Abstract
Driver fatigue has a direct impact on urban railway transit (URT) drivers’ driving behavior and can cause driver error. The existing methods for fatigue detection mainly train the models with supervised learning, relying heavily on the annotation of recorded data. However, labeled data are unobtainable in some environments, especially for URT driver fatigue levels during actual driving. Therefore, this study proposes a fatigue detection method using unlabeled heart rate variability data to monitor URT driver fatigue in actual working conditions. By utilizing the existing conclusions with regard to factors contributing to fatigue and physiological changes, this study annotated a small number of samples and then used a novel positive and unlabeled learning algorithm based on nearest neighbors and random forest to divide samples into different fatigue levels. The proposed method was evaluated using the URT driver fatigue data sets collected in the lab. Binary classification achieved an accuracy of 79.0%. However, the accuracy of three-class classification was only 55.7%. In addition, the proposed method performed as well using the field data set as it did using the lab data set. The results show the high generalization performance of the proposed method, which could contribute to addressing the issue of lack of labeled training data for fatigue detection in actual working conditions.
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 | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 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".