作者: H. Ahmadi , A. Golian , M. Mottaghitalab , N. Nariman-Zadeh
关键词:
摘要: A group method of data handling-type neural network (GMDH-type NN) with an evolutionary genetic algorithm was used to predict the TME n feather meal (FM) and poultry offal (POM) based on their CP, ether extract, ash content. Thirty-seven lines consisting 15 FM 22 POM samples were collected from literature train a GMDH-type NN model. deployed design whole architecture NN. The accuracy model examined by R 2 value, adjusted , mean square error, residual standard deviation, absolute percentage bias. developed could accurately or chemical composition. for had higher prediction than models reported previously. This study revealed that novel modeling can be by-products.