scientific article; zbMATH DE number 4012303

From MaRDI portal
Publication:3762062





The use of nondominated extreme points generating algorithms on multicriteria linear programming raises several problems. These difficulties lead us to focus our attention on the development of the interactive methods. In this paper we describe a new interactive method we have developed. This method is based on a progressive and selective learning of the Pareto optimal solutions set. The intention is not to converge into the best compromise solution but to help the decision maker to eliminate the subsets of solutions which are of no interest to him. This reduces the calculations and leads to a limited number of solutions for final selection. There are no irrevocable decisions along the process being always admissible to go backwards. The method is applicable to problems with three objective functions. Although being a limitation this allows the use of graphical means specially convenient for the dialogue with the decision maker. It should be noticed that Pareto optimal points may be analysed which are not vertices of the feasible polyhedron. Finally we will refer that the method is not too demanding with respect to the information required from the decision maker in each iteration, the main concern being to have a simple and flexible procedure.





Describes a project that uses

Uses Software






This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3762062)