PRIORITY VECTOR ESTIMATION: CONSISTENCY, COMPATIBILITY, PRECISION

Abstract

Various methods of priority vector estimation are known in the Analytic Hierarchy Process (AHP). They include the classical eigenproblem method given by Thomas Saaty, developments in least squares and the multiplicative approach, robust estimation based on transformation of the pairwise ratios to the shares of preferences, and other approaches. In this paper, the priority vectors are completed with validation of data consistency, comparisons of vectors' compatibility, and estimation of precision for matrix approximation by vectors. The numerical results for different data sizes and consistency show that the considered methods reveal useful features, are simple and convenient, and capable of facilitating practical applications of the AHP in solving various multiple-criteria decision making problems.