Use of linear modeling, multivariate adaptive regression splines and decision trees in body weight prediction in goats

Use of robust regression algorithms for better prediction body weight (BW) is receiving increased attention. The present study therefore aimed at predicting BW from chest circumference, breed and sex a total 1,012 goats. animals comprised 332 matured West African Dwarf (WAD) (197 bucks 135 does), 374 Red Sokoto (RS) (216 158 does) 306 Sahel (SH) (172 134 randomly selected in Nasarawa State, north central Nigeria. was made using automatic linear modeling (ALM), multivariate adaptive splines (MARS), classification tree (CART), chi-square interaction detection (CHAID) exhaustive CHAID. predictive ability each statistical approach measured goodness fit criteria i.e. Pearson?s correlation coefficient (r), Coefficient determination (R2), Adjusted (Adj. R2), Root-mean-square error (RMSE), Mean absolute percentage (MAPE), deviation (MAD), Global relative approximation (RAE), Standard ratio (SD ratio), Akaike?s information criterion (AIC) corrected (AICc). Male RS SH goats had significantly (P<0.05) higher CC compared to their female counterparts while WAD, male (57.88?0.51 vs. 55.45?0.55). determined be the trait paramount importance prediction, as expected. Among five models, MARS algorithm gave best with r, R2, Adj. SDratio, RMSE, RAE, MAPE, MAD, AIC AICc values 0.966, 0.933, 0.932, 0.26, 1.078, 0.045, 3.245, 0.743, 186.0 187.0, respectively. may guide choice model which exploited selection genetic improvement including feed health management marketing purposes, especially identification studied breed?s standards.