Enhancing project selection: elicitation of parameter values for effective multiple criteria project sorting

Effective decision-making is crucial to the success of an organisation's project selection problem as it involves addressing complex challenges. One such challenge, known as multiple criteria sorting, is to assign projects into ordered classes considering the preferences of a decision maker (DM). The authors recently proposed a comprehensive method to address highly complex problems based on the outranking approach. However, it is widely acknowledged that determining parameter values for methods based on the outranking approach can be difficult. This difficulty stems from the large number of parameters involved and the DM's limited familiarity with them. Here, we focus on two aspects: 1) how to obtain the parameter values for the method; 2) how to incorporate imperfect knowledge during the sorting process. We adopt the preference disaggregation paradigm and employ evolutionary algorithms. Our proposed approach demonstrates excellent performance in a wide range of computational experiments.