Equivalent Exchange Method for Decision-Making in Case of Alternatives with Incomparable Attributes

Abstract

The paper is focused on searching for novel methods aimed at improving the performance and usability of a common decision-making process where a panel of experts are assisted by specialized software systems. An equivalent exchange method (EEM) is considered in the paper as a potential candidate for a versatile method applicable in expert decision-making process for solving problems in various subject domains. The method is formally described in the paper in the form of an iterative algorithm where each iteration leads to the reduce in the number of alternatives under consideration until it converges to the preferable one. The key feature of EEM consists in the fact that the original comparison between multiple alternatives described by many attributes measured in different units is replaced by the sequence of simple exchanges between pairs of alternatives where only two attributes are engaged at once. The numerical example illustrating the full run of the algorithm is thoroughly described, so the actions performed in the steps of the algorithm are explained. The case of the successful implementation of EEM as the module of Expert Decision Support System is also presented.

Keywords

• decision support systems
• decision-making
• preference matrix
• comparison matrix
• equivalent exchange method
• dominated alternatives