The objective of this study was to compare predictive performances of four machine learning (ML) models: Support Vector Machines with Radial Basis Function Kernel (SVMR), Classification and Regression Trees (CART), Random Forest (RF) and Model Average Neural Networks (MANN) to predict live weight from morphological measurements of Norduz sheep (n=93). Seven morphological measurements; chest girth (CG), chest width (CW), chest depth (CD), height at withers (HW), body length (BL), heigth at rump (HR) and rump width (RW) were used to predict live weigth (LW) of Norduz sheep. All morphological measurements were positively correlated to LW. Live weight had the highest correlation with CG and the lowest correlation with HR. Initially, highly correlated predictors were removed from the data set. The remaining predictors were then subjected to variable selection procedures using the Boruta algorithm. The results of Boruta confirmed the importance of the four predictors HW, BL, CW, and CD. However, HR confirmed to be unimportant was excluded from the dataset. The ML models were trained on selected predictors. The results showed that the prediction performance validated using the test dataset indicated that RF had the lowest values of Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Mean Absolute Percent Error (MAPE). The permutation-based variable importance scores indicate that CW and CD were the most important variables in predicting LW. The actual LW had the highest significant positive correlations with the values predicted by SVMR and RF, and followed by ANN and CART models respectively. There were no differences between the means of actual and predicted LWs by machine learning models. The fact that the models generalized well on the testing data sets indicates that machine learning algorithms have valid predictive patterns and are effective methods in LW weight of Norduz sheep. Considering runtime of the models, although the CART model had the lowest computational cost, it had the worst performance. The MANN algorithm, on the other hand, required a longer runtime to process the same dataset.

Measurements of the body structure in sheep are worthy of judging the quantitative features of meat and useful in developing appropriate selection requirements. The current study was aimed to predict live weight and hot carcass weight from linear body measurements of yearling male local sheep. 84 days feeding period fortnightly taken data on 24 local sheep for body weight, body length, heart girth, wither height, sub-sternal height, tail length, tail width, scrotal circumference, and scrotal length were analyzed to study the relationship between linear body measurements and body weight. At the end of the trial all sheep were slaughtered to measure the relationship between body measurements, and hot carcass weight. Microsoft Excel 2010 was used for data analysis. The relationships between the various body measurements were calculated using pearson's correlation coefficient. The backward stepwise multiple regression procedure was used for the determination of the most suitable model for the prediction of the live weight and hot carcass weight. Hot carcass weight was highly correlated (P<0.01) with body weight and scrotal circumference. Besides, it was significantly (P<0.05) correlated with tail width. Body weight was significantly (P<0.05) correlated with all body measurements except tail length and scrotal length. It is concluded that the body weight of the local sheep can be predicted with heart girth, sub-sternal height and tail width; the equation is LW= -97.2 + 0.36HG + 2.1SBSH + 0.57TW with a better coefficient of determination; R2 = 0.55 and the hot carcass weight can be predicted with sub-sternal height and tail width; the equation is HCW= -75.66 + 1.75SBSH + 0.85TW with a coefficient of determination; R2 = 0.33. But, hot carcass can be predicted with body weight, the equation is HCW= -9.39+0.85BWT when weighing scales are affordable with a better coefficient of determination; R2= 0.557.