Braeden Fieguth's MSc Defence

Applying machine learning to predict future milk production in dairy cows

This thesis comprises three main chapters. First, a literature review explores fundamental machine learning concepts, common linear regression algorithms and optimization techniques. Second, it introduces "Brisk," a Python package developed to make machine learning more accessible for non-computer science users, detailing its motivations, usage, and a practical case study. Finally, the thesis presents two machine learning models developed for a decision support system. These models predict day-305 milk yield at 60 days in milk. These models aim to identify cows suitable for an extended lactation management strategy.