Comparison of responses to dietary protein and lysine in broiler chicks reared before and after 2000 via neural network models

Abstract

SUMMARYThe current study was conducted to compare the responses of broiler chicks (average daily gain (ADG) and feed efficiency (FE)) raised before and after 2000 to dietary protein and lysine through neural networks (NN). The available lysine dose-response data were extracted from the literature and arbitrarily divided into two sets of before and after 2000. The training and testing data sets derived from each group were used to develop the NN models. The developed models were subjected to a sensitivity analysis test to assess the relative importance of dietary protein and lysine on chicks' responses. An optimization algorithm was used to find the dietary protein and lysine required for maximum ADG and FE, based on each dataset. The results showed that the NN models developed could predict ADG and FE efficiently in broiler chicks of before and after 2000, and the higher accuracies of prediction were attained by these models compared to those of regression models. Sensitivity analysis indicated that ADG and FE were more sensitive to dietary lysine, compared to protein, in both time periods. Based on the optimization results, the protein and lysine requirements for maximum ADG or FE for birds reared after 2000 were lower and higher, respectively, compared to those reared before 2000. The protein requirements for maximum ADG and FE for birds reared before 2000 were 241·3 and 247·0 g/kg diet and for lysine 10·76 and 11·18 g/kg diet, respectively. In birds reared after 2000, maximum ADG was obtained when the diet contained 224·30 g protein/ kg diet and 11·75 g lysine/kg diet, whereas maximum FE was achieved with a diet containing 228·3 g protein and 13·1 g lysine.

Effect of protein concentrations in the diet on productive performance, carcass characteristics, and meat chemical composition of broiler chickens in the dry subtropics

“…ISSN 2007 -0705, pp. : 1 -18 -4 -24.1% (Faridi et al, 2012) and 24.5% (Faridi et al, 2015). In another study, Abdel-Maksoud et al (2010) reported maximum body weight in starter broiler chickens with lower dietary CP levels.…”

Predicting the metabolizable energy content of corn for ducks: a comparison of support vector regression with other methods

“…The network randomization was set to normal with mean and variance of 0 and 0.1, respectively. This option specifies how the weights should be initialized at the beginning of the calibration process (Faridi et al, 2012b). Statistica Neural Networks version 8.0 software was used to construct and calibrate the ANN models (StatSoft, 2009).…”