Optimizing Grass Selection for Beef Cattle Using Multi-Armed Bandit Algorithms: A Data-Driven Approach to Enhance Growth Through Rumination Analysis

This paper explores the application of the Multi-Armed Bandit (MAB) approach to optimize the selection of grass types for cattle to enhance their growth. Utilizing data from a cow monitoring system, CowManagement, specifically rumination, which is a strong indicator of nutrient absorption, the study effectively evaluates different grass options. The cattle are fed with a single type of grass and are allowed to switch to a different grass type at the same interval while the CowManager sensors continuously collect the rumination data. The Upper Confidence Bound (UCB) algorithm is used to identify the optimal grass type. This method provides a data-driven framework for improving cattle growth. The results confirm the effectiveness of the MAB strategy in optimizing grass selection, offering valuable insights for sustainable and efficient livestock farming.