abstract
- There are several difficulties with GP. This paper is an appraisement to the effectiveness of GP in incorporating DM’s preferences and explains some fundamental relationships among the different approaches of incorporating. Since each approach has its own way and properties, we integrate the approaches by a unique classification. We explain why and how inappropriate approximation of nonlinear functions may lead to missing information about DM’s preferences and how we can minimize the lost information in GP. Finally, an effective and interactive procedure with a numerical example is suggested in order to help analysts in defining the DM’s preference functions more effectively.