Application of mathematical models for true metabolisable energy in sorghum grain for poultry

Abstract

Sorghum grain is an important ingredient in poultry diets. Nitrogen-corrected true metabolizable energy (TMEn) content of sorghum grain is a measure of its quality. As for the other feed ingredients, the biological procedure used to determine the TMEn value of sorghum grain is costly and time-consuming. Therefore, it is necessary to find an alternative method to accurately estimate the TMEn content <br><br>of sorghum grain. Artificial neural networks are the powerful method which widely used in agriculture and poultry nutrition. Therefore In this study, an artificial neural network (ANN) and a multiple linear <br><br>regression (MLR) models were used to predict the TMEn of sorghum grain based on its acid detergent fiber (ADF) and total phenols content. The accuracy of the models was calculated by R2, MS error and bias. The predictive ability of an ANN was compared with a MLR model using the same training data sets. The results of this study showed that it is possible to estimate sorghum grain TMEn with a simple analyti-<br><br>cal determination of ADF and phenolic content. The R2 values corresponding to testing and training of the ANN model showed a higher accuracy of prediction than that established by regression method (R2 <br><br>= 0.84 vs 0.56 for training and R2 = 0.83 vs 0.47 for testing data sets respectively). In conclusion, the ANN model may be used to accurately estimate the TMEn value of sorghum grain from its correspond-<br><br>ing chemical composition (ADF and total phenols content)