Evaluating the effect of salinity on corn grain yield using multilayer perceptron neural networks

Author's Name: O. M. Ibrahim
Subject Area: Life Sciences
Subject Agricultural
Section Research Paper


salinity, relative importance, multilayer perceptron neural network


A multilayer perceptron neural network (MLP) is a powerful statistical modeling technique in the agricultural sciences. The aim of the relative importance analysis is to separate explained variance among multiple predictors to better understand the role played by each predictor or independent variable. To assess the relative importance of independent variables especially which have negative relationship with the dependent variable in multilayer perceptron neural network the current study was carried out to make a comparison among different algorithms, the Connection Weights Algorithm, Modified Connection Weights Algorithm, Most Squares Algorithm, Dominance Analysis, Garson’s Algorithm, Partial Derivatives and Multiple Linear Regression. The performance of these algorithms is studied for empirical data. The Most Squares Algorithm is found to be a better Algorithm in comparison to the above mentioned Algorithms and seem to perform much better, and agree with the results of multiple linear regressions in terms of the partial R2 and seem to be more reliable.

Download Full Paper