A sampling-search-clustering approach for exploring the feasible/efficient solutions of MCDM problems

Abstract Decision making under certainty of a single decision maker (DM) is considered. The interaction between the DM and a computerized decision aid (DA) is discussed. The DA can submit feasible, efficient or optimal solutions depending on the formal problem discription provided by the DM. It is proposed that the definition of optimal solution for MCDM problems should be based on the DM's confidence that this solution has been obtained. Methods based on the confidence concept are designed. The suitability of the presented approach for tackling non-linear, non-convex and multi-modal MCDM problems is demonstrated with the aid of two sample problems.