Predicting body weight of Kalahari Red goats from linear body measurements using data mining algorithms

Abstract

Background and Aim: The Kalahari Red goat breed is the finest meat-producing species in South Africa, and its coat color ranges from light to dark red-brown. A practical approach to estimating their body weight (BW) using linear body measurements is still scarce. Therefore, this study aimed to determine the best data mining technique among classification and regression trees (CART), Chi-square automatic interaction detection (CHAID), and exhaustive CHAID (Ex-CHAID) for predicting the BW of Kalahari Red goats. Materials and Methods: This study included 50 Kalahari Red goats (does = 42 and bucks = 8) aged 3–5 years. Body length (BL), heart girth (HG), rump height (RH), height at withers (WH), sex, and age were the essential indicators to estimate BW. The best model was chosen based on the goodness of fit, such as adjusted coefficient of determination (Adj. R2), coefficient of determination (R2), root mean square error (RMSE), standard deviation ratio (SD ratio), mean absolute percentage error, Akaike information criteria, relative approximation error, and coefficient of variation. Results: The SD values of the ratio ranged from 0.32 (CART) to 0.40 (Ex-CHAID). The greatest R2 (%) was established for CART (89.23), followed by CHAID (81.99), and the lowest was established for Ex-CHAID (81.70). CART was established as the preferred algorithm with BL, HG, and WH as critical predictors. The heaviest BW (73.50 kg) was established in four goats with BL higher than 92.5 cm. Conclusion: This study reveals that CART is the optimum model with BL, HG, and WH as the essential linear body measurements for estimating BW for Kalahari Red goats. The updated records will assist the rural farmers in making precise judgments for various objectives, such as marketing, breeding, feeding, and veterinary services in remote areas where weighing scales are unavailable.