The Use of Selected Machine Learning Methods in Dairy Cattle Farming: A Review
Wilhelm Grzesiak, Daniel Zaborski, Marcin Pluciński, Magdalena Jędrzejczak-Silicka, Renata Pilarczyk, Piotr Sablik
Abstract
The aim of this review was to present selected machine learning (ML) algorithms used in dairy cattle farming in recent years (2020-2024). A description of ML methods (linear and logistic regression, classification and regression trees, chi-squared automatic interaction detection, random forest, AdaBoost, support vector machines, k-nearest neighbors, naive Bayes classifier, multivariate adaptive regression splines, artificial neural networks, including deep neural networks and convolutional neural networks, as well as Gaussian mixture models and cluster analysis), with some examples of their application in various aspects of dairy cattle breeding and husbandry, is provided. In addition, the stages of model construction and implementation, as well as the performance indicators for regression and classification models, are described. Finally, time trends in the popularity of ML methods in dairy cattle farming are briefly discussed.