Identification of Weights in Multi-Criteria Decision Problems Based on Stochastic Optimization.

Many scientific papers are devoted to solving multi-criteria problems using methods that find discrete solutions. However, the main challenge addressed by our work is the case when new decision-making variants have emerged which have not been assessed. Unfortunately, discrete identification makes it impossible to determine the preferences for new alternatives if we do not know the whole set of parameters, such as criteria weights. This paper proposes a new approach to identifying a multi-criteria decision model to address this challenge. The novelty of this work is using a discretization in the space of the problem to identify a continuous decisional model. We present a hybrid approach where the new alternative can be assessed based on stochastic optimization and the TOPSIS technique. The stochastic methods are used to find criteria weights used in the TOPSIS method. In that way, we get assessed easily each new alternative based only on the initial set of evaluated alternatives. [ABSTRACT FROM AUTHOR]