New hybrid EC-PROMETHEE method with multiple iterations of random weight ranges: Step-by-step application in Python

The decision-making process consists of finding the best solution to an analyzed problem. This search is carried out in the face of countless interactions when analyzing an alternative criterion by criterion, under which weights are assigned that distinguish the degree of importance they have for the decision-makers. The definition of weight for each criterion gives rise to three lines of thought on the subject. There are objective, subjective, and hybrid methods. This discussion concerns the degree to which experts define the criteria weights. Based on this discussion, we developed a hybrid method to integrate the Entropy and CRITIC methods with the PROMETHEE method, called EC-PROMETHEE. The innovation of this method is that the combination of the Entropy and CRITIC methods does not result in a single set of weights. In reality, the weights generated by each method are used to define each criterion's upper and lower limits. The range of weights generated for each criterion is emulated "n" times and builds a set of weights that are applied to the ranking definition process. The model generates "n" rankings, defining a single ranking. In this article, we demonstrate a step-by-step application of a tool developed in Python called EC-PROMETHEE and use it as an example of the problem of choosing rotary-wing airplanes for application in the military police service.

• ➢

  The method reduces discretion in determining the weights of the criteria;

• ➢

  The innovation lies in the use of a range of weights for criteria;

• ➢

  Consistency in defining the final ranking.