Use of machine-learning and visualization techniques in the evaluation of factors affecting Milk Urea Nitrogen
Bibliographic record
Abstract
Studies suggest that milk urea nitrogen is influenced by multiple dietary and non-dietary factors; however most studies continue evaluating those effects independently. Further information is required in order to understand the properties, variations, and applicability of milk-nitrogen fractions by the producers. The objective of this study was to use machine-learning and visualization techniques in the investigation and evaluation of multiple factors altering milk urea nitrogen. Records from the Quebec Dairy Production Centre of Expertise (Valacta) were used in the analyses. After edits, the data consisted of 2,382,043 milk test-day and feeding composition records from Ayrshire, Brown Swiss, Holstein, and Jersey cows. Mean milk urea nitrogen varied across breeds (12.13 ± 3.71 mg/dL; 13.52 ± 3.82 mg/dL; 11.1 ± 3.43 mg/dL; and 13.78 ± 3.8 mg/dL in Ayrshire, Brown Swiss, Holstein, and Jersey, respectively) and across lactation (milk urea nitrogen concentrations increased with parity number). Decision-trees were generated to determine the attributes associated with milk urea nitrogen levels. Results indicated that the most significant variables altering milk urea nitrogen were milk-fat percentage, dietary crude protein, herd size, and somatic cell count. Milk-fat percentage and dietary crude protein appeared to interact with milk urea nitrogen over the entire lactation. Visualization techniques aided in the identification of changes in feeding practices. During early stages of lactation, producers tended to offer diets with high crude protein content. During medium and late stages of lactation, producers seemed to over-feed their cows, producing an increase in milk urea excretion. Apart from sub-optimal management practices, these results also point to higher feeding costs as well as potential increases in environmental emissions of nitrogen and ammonia.
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 | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 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 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".