Early Identification of Individual Nursery Pigs at Risk of Requiring Health Treatment Using Machine Learning
Bibliographic record
Abstract
Infectious disease is a major concern in the swine industry, impacting production as well as animal welfare. In this research, a prediction model for early identification of late nursery pigs within a batch that are at higher risk of requiring health treatments in a dynamic disease environment was developed based on early daily data on feeding and drinking and body weight. For this purpose, a unique dataset was used consisting of 21 batches of up to 75 late nursery pigs that were entered into a natural polymicrobial disease challenge barn to develop and evaluate the prediction models. The model was designed to predict the probability of a pig requiring at least one health treatment between 14 and 27 days following exposure to the disease challenge based on daily feeding, drinking, and body weight data collected on the pig and the batch from days 6 to 14 after exposure. Four tree-based machine-learning models and an ensemble model were used to develop the prediction models using the leave-one-batch-out approach for training and validation. The prediction results were further used to rank pigs within a batch on the predicted probability of requiring treatment. All models were evaluated in terms of area under the curve (AUC), accuracy, and Pearson correlation between predictions and observed outcomes (treated or not). In general, all models had a limited ability to predict the number of pigs that required at least one treatment for a new batch because of the dynamic nature of the disease challenge between batches and the use of batch-level medications and other interventions. However, all models had some ability to rank pigs based on the probability of requiring treatment and these probabilities were generally positively correlated with outcomes (treated or not treated between day 14 and 27 after exposure) within a batch, although these correlations were highly variable between batches, ranging from -0.13 to +0.48, and averaged around 0.22. All models had similar prediction performance, although Random Forest generally had the highest performance. In general, we conclude that early daily data on feeding, drinking, and body weight has some ability to identify nursery pigs within a batch that are at higher risk of requiring health treatments but data on additional features or human observations will be needed to improve early identification of such pigs. Finally, drinking data provided slightly more information than feeding data.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 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".