A Probabilistic Approach to Multiple Attribute Decision Making with Fuzzy Preference Information on Alternatives.

Fuzzy Multiple Attribute Decision Making (MADM) problems are commonly encountered in everyday aspect of life. They aim at selecting the optimal alternative among some courses of action in the presence of multiple, usually conflicting, attributes with fuzzy performance ratings. The selection procedures of ranking alternatives of almost all the studies in the area of fuzzy MADM are based on the decision matrix. However, in real situations, the selection procedure of an MADM problem is usually conducted by a group of experts (or consultants). Due to the intrinsic differences between these experts (or consultants), the performance ratings for each alternative of the alternatives obtained from them are different from one another. Hence, it seems to be inappropriate to use only one value as the representative of the performance ratings for each alternative of the alternatives. The main purpose of this paper is to determine the attribute weights of the MADM problem where the decision maker's preference information on alternatives is known and is described by a reciprocal binary fuzzy relation through probability techniques. Besides, after the attribute weights being determined, we further concern with the problem of determining the optimal number of the performance ratings for each attribute of the alternatives. [ABSTRACT FROM AUTHOR]