Evaluation of feed efficiency in beef cattle using artificial intelligence and metabolite profile

• Grant number: 23/11281-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: December 01, 2023
• End date: March 31, 2025
• Field of knowledge: Agronomical Sciences - Agronomy
• Principal Investigator: Rafael Vieira de Sousa
• Grantee: Lucas Basolli Borsatto
• Host Institution: Faculdade de Zootecnia e Engenharia de Alimentos (FZEA). Universidade de São Paulo (USP). Pirassununga , SP, Brazil
• Associated scholarship(s): 24/02625-2 - Enhancing feed efficiency prediction in beef cattle through metabolite profile and artificial intelligence techniques, BE.EP.IC

Abstract

The methods employed in predicting feed efficiency (FE) in beef cattle can play a crucial role in improving feed management and selecting more efficient animals. The aim of this study is to evaluate artificial intelligence methods including supervised and unsupervised machine learning algorithms to classify FE in beef cattle using blood metabolite data. For this purpose, an existing raw database will be used, which will be preprocessed and organized for the modeling stage. This database consists of information on metabolites obtained from 64 Black Angus cattle. The modeling process consists of two distinct stages: the first is dedicated to selecting subsets of metabolites that show the highest correlation with FE, and the second stage involves training and comparing different models for predicting feed efficiency. Both stages will be developed using the Weka 3.8.6 software for preliminary modeling evaluations (quick simulation), and the Python language with machine learning libraries for building and evaluating the final models (refinement).