Estimation of Cold Carcass Weight and Body Weight from Several Body Measurements in Sheep through Various Data Mining Algorithms

The goal of the present study was to compare the predictive performance of three data mining algorithms viz., CHAID, Exhaustive CHAID, and CART implemented in the estimation of cold carcass weight (CCW) and body weight (BW) from several body measurements (withers height (WH), chest depth (CD), body length (BL), hearth girth (HG) and leg circumference (LC)) measured from five sheep breeds (Akkaraman (9), Daglic (10), Kivircik (10), Merinos (10) and Karacabey Merino (8)) reared in Konya province conditions located in the Central Anatolia Region of Turkey. For measuring the predictive performance of three algorithms in Models I and II, goodness of fit criteria (coefficient of determination (R-2%), adjusted coefficient of determination (Adj.R-2%), coefficient of variation (CV%), SD ratio, Root Mean Square Error (RMSE), Relative Approximation Error (RAE), and Pearson correlation coefficient between actual and predicted values were calculated. For both Models, CHAID and CART were chosen as the best algorithms in the estimation of CCW trait, whereas only CHAID was the ideal tree-based algorithm in the estimation of BW trait. In conclusion, the determination of the best data mining algorithm on the estimation of BW and CCW traits might be utility for further researches linked with characterization of sheep breeds, and sheep breeding in very large flocks.