Fig. 2From: Integrating on-farm and genomic information improves the predictive ability of milk infrared prediction of blood indicators of metabolic disorders in dairy cowsRelationship between relative difference (expressed as the difference in predictive ability from Model 4, i.e., full model including FTIR data, DIM, parity and genomic information and from Model 1 considering only FTIR data) and heritability estimates for each trait. Plots are differentiated for tenfold, batch-out and herd-out cross-validation scenarios using BayesB (a, c, e) and gradient boosting machine (b, d, f) approaches for predicted blood metabolitesBack to article page