THE R-AND-D PROJECT SELECTION PROBLEM WITH FUZZY COEFFICIENTS

This paper deals with the multicriteria project selection problem subject to linear constraints, with zero-one decision variables. Linguistic rather than numerical values are used in the project appraisals for each criterion, with their meanings represented by normal fuzzy sets. The grade of membership of a given alternative, in the fuzzy set of all nondominated alternatives, is computed by an algorithm that converges rapidly on the final solution. Continuity and normality are the only restrictions imposed on the membership functions. An example of the overall procedure and topics for future research are presented.