Prediction of a quality index of bovine carcasses for elite Chilean market using a mathematical model

Abstract

The Chilean beef sector has to satisfy the international demand of beef with homogenous and high quality products. Since quality is a subjective parameter, it is necessary to find the variables to objectively characterize the carcasses. Further development of mathematical models as tools for better prediction and classification of beef quality is needed. The aim of this study was to predict beef quality for the Chilean market, using the data generated by the programme “Programa de Desarrollo de Proveedores (PDP)” implemented by Carnes Ñuble S.A. between 1999 and 2003. Twelve variables were generated. Five of them received a score and its addition resulted in the quality index. This index allowed the classification of the meats as “common” and “special”, being the last group the one used by the “Pampaverde”^® brand. The data were studied through a descriptive analysis, as well as correlation analysis, considering the variables that contributed more to the total score. Moreover, a multiple lineal regression analysis was conducted. Using this information, a model was built to predict Chilean beef quality. The model was evaluated and validated using data from 2002.