Using Machine Learning to Classify Lightweight Pigs
Author Information
Author(s): Pau Salgado-López, Joaquim Casellas, Iara Del Pilar Solar Diaz, Thomas Rathje, Josep Gasa, David Solà-Oriol
Primary Institution: Autonomous University of Barcelona
Hypothesis
Can machine learning methods effectively classify pigs at risk of growth retardation in commercial conditions?
Conclusion
Machine learning algorithms can accurately identify pigs at risk of growth retardation, improving management and efficiency in pig production.
Supporting Evidence
- Random forest and generalized boosted regression models showed superior classification performance with AUC estimates ranging from 0.772 to 0.861.
- The study analyzed data from pigs at different production stages to identify those at risk of growth retardation.
- Key prediction factors varied across production stages, highlighting the importance of birthweight and weight at previous stages.
Takeaway
This study shows that computers can help farmers find pigs that are growing too slowly, so they can take better care of them.
Methodology
Data from 26,749 pigs was analyzed using machine learning algorithms and ordinary least squares to classify pigs based on weight at different production stages.
Limitations
The study was conducted on data from a single commercial farm, which may limit the generalizability of the findings.
Participant Demographics
Crossbred pigs (Yorkshire × Landrace with Duroc) from a commercial breeding herd.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website