Search

Advances in technology and improved data collection have increased the availability of genomic estimated breeding values (gEBV) and phenotypic information on dairy farms. This information could be used for the prediction of complex traits such as survival, which can in turn be used in replacement heifer management. In this study, we investigated which gEBV and phenotypic variables are of use in the prediction of survival. Survival was defined as survival to second lactation, plus 2 wk, a binary trait. A data set was obtained of 6,847 heifers that were all genotyped at birth. Each heifer had 50 gEBV and up to 62 phenotypic variables that became gradually available over time. Stepwise variable selection on 70% of the data was used to create multiple regression models to predict survival with data available at 5 decision moments: distinct points in the life of a heifer at which new phenotypic information becomes available. The remaining 30% of the data were kept apart to investigate predictive performance of the models on independent data. A combination of gEBV and phenotypic variables always resulted in the model with the highest Akaike information criterion value. The gEBV selected were longevity, feet and leg score, exterior score, udder score, and udder health score. Phenotypic variables on fertility, age at first calving, and milk quantity were important once available. It was impossible to predict individual survival accurately, but the mean predicted probability of survival of the surviving heifers was always higher than the mean predicted probability of the nonsurviving group (difference ranged from 0.014 to 0.028). The model obtained 2.0 to 3.0% more surviving heifers when the highest scoring 50% of heifers were selected compared with randomly selected heifers. Combining phenotypic information and gEBV always resulted in the highest scoring models for the prediction of survival, and especially improved early predictive performance. By selecting the heifers with the highest predicted probability of survival, increased survival could be realized at the population level in practice. [ABSTRACT FROM AUTHOR]

In this study, we compared multiple logistic regression, a linear method, to naive Bayes and random forest, 2 nonlinear machine-learning methods. We used all 3 methods to predict individual survival to second lactation in dairy heifers. The data set used for prediction contained 6,847 heifers born between January 2012 and June 2013, and had known survival outcomes. Each animal had 50 genomic estimated breeding values available at birth and up to 65 phenotypic variables that accumulated over time. Survival was predicted at 5 moments in life: at birth, at 18 mo, at first calving, at 6 wk after first calving, and at 200 d after first calving. The data sets were randomly split into 70% training and 30% testing sets to evaluate model performance for 20-fold validation. The methods were compared for accuracy, sensitivity, specificity, area under the curve (AUC) value, contrasts between groups for the prediction outcomes, and increase in surviving animals in a practical scenario. At birth and 18 mo, all methods had overlapping performance; no method significantly outperformed the other. At first calving, 6 wk after first calving, and 200 d after first calving, random forest and naive Bayes had overlapping performance, and both machine-learning methods outperformed multiple logistic regression. Overall, naive Bayes has the highest average AUC at all decision points up to 200 d after first calving. Random forest had the highest AUC at 200 d after first calving. All methods obtained similar increases in survival in the practical scenario. Despite this, the methods appeared to predict the survival of individual heifers differently. All methods improved over time, but the changes in mean model outcomes for surviving and non-surviving animals differed by method. Furthermore, the correlations of individual predictions between methods ranged from r = 0.417 to r = 0.700; the lowest correlations were at first calving for all methods. In short, all 3 methods were able to predict survival at a population level, because all methods improved survival in a practical scenario. However, depending on the method used, predictions for individual animals were quite different between methods. [ABSTRACT FROM AUTHOR]