The contribution of ANN's simple perceptron pattern to inequalities measurement in Regional Science

Abstract

The present paper presents a synthetic approach to the XOR operation structure combining the cognitive matter of different fields of scientific study in order to elect further utility of this operation for the inequalities measurement in Regional Science. The observation of structural similarities, between the simple perceptron pattern, in artificial neural networks (ANN), and an XOR binary logic gate, in Digital Electronics Theory authorizes these two models to be considered identical for inequalities detection, under the presumption that both of them can provide a solution for the XOR problem. The solution of the XOR problem for a simple perceptron pattern introduces a new inequalities index, aroused as a generalization of the XOR problem's solution for the ANN simple perceptron model. The new index appears to operate satisfactorily in Regional Science's inequalities research, providing a significant similar behavior in comparison with the Theil index and performing better in cases that the Theil index calculation presupposes transformation treatment.