Analyzing AHP matrices by robust partial least squares regression

The Analytic Hierarchy Process (AHP) [8] is a powerful process to help people to express priorities and make the best decision when both qualitative and quantitative aspects of a decision need to be considered. In this paper, in order to eliminate the influence of outliers, we use an approach based on Robust Partial Least Squares (R-PLS)[12] regression for the computation of the values for the weights of a comparison matrix. A simulation study to compare the results with other methods for computing the weights is proposed to analyze comparison matrix.