Artificial Intelligence and Classical Methods in Animal Genetics and Breeding
摘要
Abstract—
Basic methods of population genetics and animal breeding and mathematical methods of machine learning used in animal breeding are analyzed. CatBoost library models were trained on the example of two domesticated species—horse (Equus caballus) and reindeer (Rangifer tarandus). Data from microsatellite panels of loci 16 and 17, respectively, were used to train the model using data on domesticated and wild reindeer, European and Russian horse breeds. The standard indicators (Accuracy, Precision, Recall, and F1) were calculated, and confusion matrices were constructed to assess the success of the model. New possibilities for identifying animal breed affiliation are shown.