42
https://pubmed.ncbi.nlm.nih.gov/38116292
This study compares four methods for classifying breed combinations in slaughter pigs using genomic data, and finds that Partial Least Squares together with Quadratic Discriminant Analysis (PLS-QDA) performs the best, and is a preferable method for the industry due to its ability to include features such as a separate class for “unknown breed combination” and informative priors for crossbreeds.