GRA-TRI: A Multicriteria Decision Aid Classification Method based on Grey Relational Analysis

Abstract

Multicriteria Decision Aiding (MCDA) paradigm has been utilized in solving classification problems, as well as ranking and selection problems. MCDA classification methods are distinguished from the other ones because they allow the decision maker to intervene the model and have no statistical constraints. In this study, a novel MCDA classification method based on Grey Relational Analysis (GRA) is presented. The method, which is named GRA-TRI, allows defining a category by reference alternative and realize assignment by considering the mathematical and geometrical similarity in GRA. In the paper, firstly, the notations and assumptions of the method have been described. Then, an illustrative example was performed step by step. By comparing GRA-TRI and ELECTRE TRI-Central, it has been showed that GRA-TRI makes consistent and reasonable assignments. Hereby, an alternative method which is simple and comprehensible, and has the advantages of grey system theory, has been presented.