A Learning Procedure to Estimate Missing Values in Fuzzy Preference Relations Based on Additive Consistency

In decision-making, information is usually provided by means of fuzzy preference relations. However, there may be cases in which experts do not have an in-depth knowledge of the problem to be solved, and thus their fuzzy preference relations may be incomplete, i.e. some values may not be given or may be missing. In this paper we present a procedure to find out the missing values of an incomplete fuzzy preference relation using the values known. We also define an expert consistency measure, based on additive consistency property. We show that our procedure to find out the missing values maintains the consistency of the original, incomplete fuzzy preference relation provided by the expert. Finally, to illustrate all this, an example of the procedure is presented.