Evaluating Rhythmic Representations in Mental Health from Wearable Devices Using the GLOBEM Datasets
Bibliographic record
Abstract
Recent studies have linked depression to biobehavioral rhythms, which are repeating cycles of physiological, psychological, social, and environmental patterns captured by wearable devices. However, their predictive potential still needs further exploration. This paper investigates whether modeling periodic patterns in longitudinal data, using cosine transformations and mathematical rhythm modeling, can help machine learning (ML) models learn hidden periodic patterns and improve the interpretability of depression prediction methods. Using the GLOBEM datasets, we extracted statistical, rhythmic, and data-driven learned features from wearable and mobile phone data. We observed that periodic features like SSA and Cosinor improve depression detection, with Transformer and Reordering models showing the most benefit from SSA, while statistical features often lead to better balanced accuracy. Lastly, we observed that features aimed at guiding models to learn hidden periodical patterns, such as the cosine of day and week, improve performance in models like Transformer. Feature importance analysis shows that statistical and periodic features, especially sleep and screen duration times or their corresponding phase and trend, consistently rank as the most influential across most iterations with different feature sets.
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 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 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) | 0.001 | 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".