Prediction of Nutrient Requirements in Beef Cattle Using Statistical and Artificial Neural Network Models

Abstract

This study focused on evaluating the nutrient requirements of beef cattle using data obtained from the Oklahoma State University Extension publication on beef cattle nutrition. The dataset included important production and nutritional variables such as body weight (BW), average daily gain (ADG), dry matter intake (DMI), total digestible nutrients (TDN), net energy for maintenance (NEm), net energy for gain (NEg), crude protein (CP), calcium (Ca), and phosphorus (P).
Initially, exploratory data analysis was carried out to understand the distribution and variability of the recorded variables. Correlation analysis was then performed to investigate the relationships between growth-related factors, particularly BW and ADG, and the corresponding nutrient requirements of beef cattle.
To estimate nutrient requirements, several predictive modelling approaches were explored, including Simple Linear Regression (SLR), Multiple Linear Regression (MLR), and Artificial Neural Network (ANN) models employing Rectified Linear Unit (ReLU) and Hyperbolic Tangent (Tanh) activation functions. The performance of these models was assessed using the coefficient of determination (R²) and Mean Squared Error (MSE).
The findings revealed that Multiple Linear Regression provided the most accurate predictions among the models considered. This suggests that the relationships between growth parameters and nutrient requirements in the dataset are largely linear. The study highlights the potential of statistical modeling as a practical tool for estimating nutrient requirements, thereby supporting more informed feeding strategies and management decisions in beef cattle production.