Neural Networks to Predict Breeding Values of Egg Production Using Phenotypic Information

Abstract Text: The objective of this study was to use MLP ANN to learn to predict the breeding values for total egg production (EBV-TEP) with the phenotypic records of traits that presented genetic correlation with the egg production. The EBV-TEP for 1,273 birds were predicted using the BLUP in a single-trait animal model that included hatch as fixed effect and additive genetic and residual as random effects. The inputs of the multilayer perceptron (MLP) neural networks were the phenotypic records of total egg production (TEP), age at first egg (AFE), body weight at 62 weeks of age (BW62) and egg weight at 40 weeks of age (EW40), measured on the birds. The results suggest that neural networks could be efficient to predict the breeding values for TEP using phenotypic measurements of the birds and the family information, without using the pedigree information.

Keywords:

breeding value,

egg production,

multilayer perceptron