Multi-criteria decision. Sorting problems. Partial information. Surrogate weights. PROMETHEE-ROC.
This research proposes the development of a multicriteria classification model with the use of partial information about the decision maker's preferences for the establishment of criteria weights. The study is based on the PROMETHEE-ROC method, which uses the Rank-Order Centroid (ROC) surrogate weights procedure to assign the criteria weight value. The sorting of alternatives by the proposed model will occur through a systematic assignment of alternatives into pre-defined categories based on decision/attribution rules. The illustration of the proposed model is presented through the municipalities sorting in the semi-arid region of Rio Grande do Norte to provide the strategies developed to combat drought. The application results were considered satisfactory for treating complex problems with multiple criteria, provided that the decision-maker involved presents non-compensatory rationality when evaluating alternatives. Moreover, the use of partial information to define the importance degree of the criteria and the systematization of the attribution of alternatives in pre-defined categories can be considered an important contribution to the literature on multicriteria sorting problems.