Applications of machine learning and genomic data to improve economic traits in dairy cattle

Abstract

Mating strategies are considered essential tools on herds under animal breeding programs,playing an important role to achieve genetic progress. The advent of Genomic Selection on the last decade, in addition to the improvements on reproduction techniques, shortened the generation interval, enhanced prediction reliabilities and selection intensity, respectively, which provided an expressive genetic gain across several industries. Those conditions revolutionized the way to select sires and dams, reflecting on the improvement of many lactational and non-lactational traits. Some negative facets, as the decrease of genetic diversity and increase of endogamy, could be observed on breeds under selection pressure. In order to optimize matings,and also evaluating the genetic diversity associated with genetic progress, artificial intelligence tools could be an alternative to the traditional linear methods. The usage of non-linear Machine Learning algorithms, and its non standard detection patterns capacity, must be applied tomatings schemes enhancement, specially its impact on economically important traits in the dairy industry. (AU)