Heuristic algorithms for aggregation of incomplete rankings in multiple criteria group decision making
Abstract
Most MCDA algorithms that deliver a ranking of alternatives as output are suitable for a single decision-maker. Our aim is to propose a general framework to make any of these methods usable in a group decision-making process. This is attained by implementing an output-oriented perspective, i.e., applying the method separately for each user and aggregating obtained outcomes into a compromise one. We focus on partial (incomplete) ranking at both the input and output of the proposed approach. This is an optimization problem and depending on the distance metric between rankings can be either discrete or continuous. Thus, various metaheuristics, dedicated algorithms, and neural network approaches are suitable to solve it. We consider the utilitarian and egalitarian perspectives oriented toward minimizing an average or a maximal distance from any input ranking. In addition, we account for the weights associated with the input rankings.