The prediction of live weight of hair goats through penalized regression methods: LASSO and adaptive LASSO

摘要

The least absolute selection and shrinkage operator (LASSO) and adaptiveLASSO methods have become a popular model in the last decade, especially fordata with a multicollinearity problem. This study was conducted to estimate thelive weight (LW) of Hair goats from biometric measurements and to selectvariables in order to reduce the model complexity by using penalizedregression methods: LASSO and adaptive LASSO for γ = 0.5 and γ = 1.The data were obtained from 132 adult goats in Honaz district of Denizliprovince. Age, gender, forehead width, ear length, head length, chest width,rump height, withers height, back height, chest depth, chest girth, and bodylength were used as explanatory variables. The adjusted coefficient ofdetermination (Radj2), root mean square error (RMSE), Akaike'sinformation criterion (AIC), Schwarz Bayesian criterion (SBC), and averagesquare error (ASE) were used in order to compare the effectiveness of themethods. It was concluded that adaptive LASSO (γ = 1) estimated the LWwith the highest accuracy for both male (Radj2 = 0.9048; RMSE  =  3.6250; AIC  =  79.2974; SBC  =  65.2633; ASE  =  7.8843)and female (Radj2 = 0.7668; RMSE  =  4.4069; AIC  =  392.5405; SBC  =  308.9888; ASE  =  18.2193) Hair goats when all the criteria were considered.