New management grading for pig farms: management grading system using pig carcass weight, back fat thickness and k-means algorithm.

Youngho Lim; Jaeyoung Kim; Gwantae Kim; Jungseok Choi

Objective: This study categorized farm management levels to improve the productivity and uniformity of pork from pigs shipped from farms. Methods: A total of 48,298 pigs were grouped (A, B, C, D group) using the k-means algorithm, carcass weight and backfat thickness. The results of the grouping were used to classify Farm Management Grades (A, B, C, D grade). Results: The proportion of primal cuts in pigs, according to the new classification method, increased from group A to group D for shoulder blade, shoulder picnic, and ham, but decreased for loin and belly. In the regression analysis of the five primal cuts (shoulder blade, shoulder picnic, loin, belly, and ham) production (kg) for each group, all regression equations showed low errors (MAE<0.7), indicating that the model can predict the production of primal cuts by group. As the Farm Management Grade decreased, the proportion of pigs in the group with large differences from the mean of carcass weight and backfat thickness of the whole pig increased. Conclusion: The results of this study confirmed the differences in primal cut traits by pig grouping and created a method to classify farms who ship non-uniform pigs. This is expected to provide indicators for improvement and supplementation to farms that ship uneven pigs, helping to enhance the production of standardized pigs at the farm level.

K-means clustering; SCAPULA; FARM management; PRODUCTION losses; REGRESSION analysis; SWINE farms

Animal Bioscience, 2025, Vol 38, Issue 2, p371

2765-0189

Academic Journal

10.5713/ab.24.0350