Authors: Juan Carlos Leyva López; Mario Araoz Medina
Addresses: Universidad de Occidente, Blvd. Lola Beltran y Blvd. Rolando Arjona Culiacan, Sinaloa, México ' Universidad de Occidente, Blvd. Lola Beltran y Blvd. Rolando Arjona, Culiacan, Sinaloa, 80020, Mexico
Abstract: This paper presents a multiobjective extension of the net flow rule for solving multicriteria ranking problems: how to rank a set of alternatives when the aggregation model of preferences is a known valued outranking relation in a decreasing order of preference. When the aggregation model of preferences is based on the outranking approach, special treatment is required, but some non-consistent situations of the explicit global model of preferences could happen. In this case, the exploitation phase could then be treated as a multiobjective optimisation problem. In this way, a number of solutions can be found that provide the decision-maker with insight into the characteristics of the problem before a final solution is chosen. We present a multiobjective evolutionary algorithm for improving the quality of a recommendation when a valued outranking relation is exploited; the performance of the algorithm is evaluated on a set of test problems. Our computational results show that the multiobjective genetic algorithm-based heuristic is capable of producing high-quality recommendations.
Keywords: multicriteria analysis; multicriteria ranking; valued outranking relations; multiobjective evolutionary algorithms; MOEAs; ranking procedures; net flow rule; NFR; genetic algorithms.
International Journal of Multicriteria Decision Making, 2013 Vol.3 No.1, pp.36 - 54
Received: 15 Jun 2012
Accepted: 15 Jun 2012
Published online: 05 Mar 2013 *