Applied Soft Computing, ( ISI ), Volume (87), Year (2020-2) , Pages (106006-106018)

Title : ( Applied improved RBF neural network model for predicting the broiler output energies )

Authors: sherwin amini , Morteza Taki , Abbas Rohani ,

Access to full-text not allowed by authors

Citation: BibTeX | EndNote

Abstract

The aim of this study is to optimize the Radial Base Function (RBF) parameters by combining the Response Surface Method (RSM) and Genetic Algorithm (GA) for modeling the output energies of broiler farms. For this purpose, data were collected from 210 broiler farms in Mazandaran province, Iran. The results showed that and (RBF parameters) have a significant effect on RBF performance (-value <0.05). The lack-of-fit in most cases was insignificant, except in the Trainlm algorithm. The results show that the RSM quadratic model can use for modeling based on and . The R2-adj and R2 indexes for RSM coefficients at the test stage by using Trainbr algorithm for broiler meat and manure outputs were 97.13, 96.76 and 98.70, 98.53 and by using Trainlm algorithm were, 87.95, 86.41 and 96.90, 96.50, respectively. The mean and standard deviation of RMSE and MAPE in both two training algorithms (Trainbr and Trainlm) for manure and broiler meat outputs were used to compare them based on 100 random data sets of k-fold cross validation method. The results showed that Trainbr has the lowest error and can use for high accuracy modeling. The results of sensitivity analysis showed that one-day chicks (), human labor (), food (), electricity (), diesel fuel () and machinery () have the highest effect on chicken energy modeling and , , , , and have the most impact on manure energy modeling.

Keywords

Broiler energy; Radial base function; Response surface method; Genetic algorithm
برای دانلود از شناسه و رمز عبور پرتال پویا استفاده کنید.

@article{paperid:1077521,
author = {Amini, Sherwin and مرتضی تاکی and Rohani, Abbas},
title = {Applied improved RBF neural network model for predicting the broiler output energies},
journal = {Applied Soft Computing},
year = {2020},
volume = {87},
month = {February},
issn = {1568-4946},
pages = {106006--106018},
numpages = {12},
keywords = {Broiler energy; Radial base function; Response surface method; Genetic algorithm},
}

[Download]

%0 Journal Article
%T Applied improved RBF neural network model for predicting the broiler output energies
%A Amini, Sherwin
%A مرتضی تاکی
%A Rohani, Abbas
%J Applied Soft Computing
%@ 1568-4946
%D 2020

[Download]