Supervised Multi-person Multi-criteria Decision-Making Methodologies under Risk and UncertaintyEngineering Applications

Resumen

In a realistic decision-making process, there are often multiple subjects in a problem. Each subject can include various criteria. In traditional decision-making, multiple experts usually tackle the problems; however, they are supposed to have all-inclusive competence in each subject. Moreover, human decisions are subjected to risk of cognitive imperfections. In other problems, rationality of decision-makers is vague in real life. Besides, the data of decision-making problems are associated with uncertainty. Therefore, for multidisciplinary decision-making problems, the most important challenges are: Segmentation is needed for experts as they may not have competence in all subjects of problem. Risk of imperfect decisions should be considered as rationality of experts is under question. Uncertainty of decision-making problems should be tackled with robust approaches to avoid loss of uncertain information. We attempt to efficiently tackle the above difficulties, as follows: Multi-person structures supervised by a director are considered for our methodologies. Subject-oriented expert segments structures are employed when multiple subjects exist in decision-making problem. We use the concept of fuzzy rationality as a cognitive aid to represent the degree of rationality of expert decisions in risky situations. New uncertain measures are introduced to reach complete uncertain computation and preserve all data. Therefore, the main objectives of this doctoral thesis are studying: (1) supervised multi-person structures with subject-oriented expert segments, (2) risk in uncertain environment, and (3) uncertainty to preserve all of data in decision-making problems with multiple criteria and alternatives. In this regard, three methodologies are proposed, as follows: An interval supervised multi-person multi-criteria decision-making methodology is presented. This methodology is without subject-oriented expert segments and has a non-risky decision-making process. The computation is based on interval distances of interval numbers and interval preference matrix. A risky fuzzy supervised multi-person multi-criteria decision-making methodology is developed supported on fuzzy-rationality-based fuzzy prospect theory. This methodology is without subjectoriented expert segments and formulated based on proposed fuzzy-rationality-based fuzzy prospect theory. The computation is based on fuzzy distances of trapezoidal fuzzy numbers and fuzzy distance matrix for extremum and ranking. A dynamic interval supervised multi-person multi-criteria decision-making methodology with subjectoriented expert segments is introduced. This methodology includes the proposed Subject-oriented Expert Segments and has a non-risky decision-making process. The parametric representation is employed for interval preference in interval optimization model. The aforementioned methodologies are derived by introducing the following decision-making models: BWM-based models for weighting process. MULTIMOORA-based models for ranking process. We use the aforementioned methodologies and models to tackle engineering problems in the area of industrial, biomedical, and energy sectors.