Spatio-temporal and class-imbalanced data analytics in healthcare
Bibliographic record
Abstract
Data analytics promise to deliver value to industries by providing historical insights that can drive the future decisions. However, depending on the characteristics of the data, machine learning models are not always accurately generalizable. For example, the well-studied problem of overfitting can arise as a result of small sample sizes. We worked on two problems in the healthcare industry, where the high dimensionality of the state space, coupled with the rarity of training samples poses challenges to the applicability of general methods to those specific problems. In the first problem, we used temporal flu activity from multiple locations in Ontario, and showed that depending on the surveillance variable under study, spatial and temporal models can each exhaust the limited amount of spatio-temporally recorded data more efficiently than the other, and predict surveillance variables more accurately as a result. In the second problem, we used a clinical dataset of radiotherapy treatment plans, whose quality is labelled by clinicians as acceptable or unacceptable, and developed an automated quality assurance system. Due to limitations in recording unacceptable plans, there is a severe class imbalance in labels that requires special learning treatments. We investigated two classification approaches, namely a class-specific learning for binary classification and a one-class learning for anomaly detection that both benefit from an adaptive resonance learning scheme that can adapt to long-term trends in labelling behaviour.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.004 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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 teacher head, 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".