Related Experiment Video
Updated: Jul 16, 2026
07:46
Exploring the Longissimus Muscle: Unraveling its Correlation with Meat Quality in Bos indicus and Crossbred Bulls
Published on: July 12, 2024
Batch-Level Average Market Weight Estimation and Interpretability Analysis for Pigs with Regional K-Means Clustering
Yan Chen1, Ruiwen Liu1, Yanbin Liang2
1College of Mathematics and Informatics, South China Agricultural University, Guangzhou 510642, China.
Animals : an Open Access Journal From MDPI
|July 15, 2026
Summary
This study introduces RC-AMFormer, a novel method using tabular data for non-invasive batch-level pig weight estimation. It improves accuracy in modern pig farming for better management and economic efficiency.
Area of Science:
- Agricultural Science
- Machine Learning
- Animal Science
Background:
- Accurate batch-level weight estimation is crucial for optimizing feeding strategies and economic efficiency in modern pig farming.
- Existing vision-based methods face limitations due to environmental factors and individual pig focus, hindering batch-level applications.
Purpose of the Study:
- To develop a non-invasive method for estimating batch-level average market weight using tabular production data.
- To explore the potential of production records for group-level pig weight assessment.
Main Methods:
- A novel approach, RC-AMFormer, was developed using over 35,000 production records.
- Key techniques include region-wise K-means clustering, cyclic month encoding, organizational hierarchy features, and an arithmetic attention mechanism.
Main Results:
- RC-AMFormer achieved a Mean Absolute Error (MAE) of 2.64 and a Mean Squared Error (MSE) of 11.77.
- The model demonstrated superior performance compared to ten other models, reducing MSE by 4.03% and MAE by 1.43% versus XGBoost.
- SHAP analysis identified key features influencing market weight.
Conclusions:
- Tabular production data offer a viable and effective approach for non-invasive, batch-level pig weight estimation.
- The findings support improved production management and economic outcomes in pig farming and similar agricultural settings.
