Prediction of live weight in Aleppo goat kids using morphometric traits: comparison of regression and machine learning models under limited sample conditions

Accurate estimation of live weight using morphometric measurements is important for herd management when direct weighing is impractical under field conditions. This study evaluated the predictive performance of regression-based and machine learning models for estimating live weight in Aleppo goat kids using a limited and relatively homogeneous dataset. A total of 75 kids were measured for withers height, rump height, body length, chest depth, chest width, chest girth, and cannon bone circumference. Ordinary Least Squares (OLS), Ridge regression, Bayesian regression, Partial Least Squares (PLS), Support Vector Regression (SVR), Artificial Neural Network (ANN), Random Forest (RF), Light Gradient Boosting Machine (LightGBM), and Categorical Boosting (CatBoost) were compared using repeated 5-fold cross-validation (10 repeats). Model performance was evaluated using RMSE, MAE, R2, RPD, and RPIQ. Ridge regression provided the strongest overall predictive performance (RMSE = 0.646; R2 = 0.799), followed by SVR and PLS. Across all models, chest girth consistently emerged as the most informative predictor of live weight. These findings suggest that, under limited and homogeneous data conditions, relatively simple morphometric measurements, particularly chest girth, combined with regularized regression approaches such as Ridge regression, may provide reliable estimates of live weight in young Aleppo goat kids raised under similar management conditions.