Models, Methods and Applications of Group Multiple-criteria Decision Analysis in Complex and Uncertain Systems

Much of existing decision making research focuses on algorithms designed to measure the preferences of different experts on a finite set of alternatives with regard to multiple conflicting criteria and trade them off alongside objective problem-relevant data. Moreover, a number of analytical methods have been developed to handle informational gaps and uncertainty. However, practical decision structures and problem requirements are versatile and more complex than the existing single-method approaches assume. Therefore, the issue of constructing user-friendly decision support procedures able to capture complexity and produce valid solutions is a relevant and unresolved problem. This thesis addresses the challenges of developing system-oriented decision support models and methods for complex multi-criteria group decision problems involving uncertainty. In essence, this research builds upon multicriteria decision analysis (MCDA) theory and system analytical concepts. First of all, the limitations and advantages of different MCDA techniques are analyzed and new integrated methodologies are developed to benefit from the strengths of individual methods while avoiding their weaknesses. Furthermore, this thesis studies the existing empirical and analytical approaches to group decision making and proposes novel methods to aggregate the opinions of different experts. In particular, the focus is on the aspects of group structuring and responsibilities definition, as well as on measuring and analyzing subjective expert estimates‟ reliability and discordance of opinions in groups. Finally, the developed methods are capable of considering informational uncertainties by using the tools of fuzzy set theory and its generalizations. Additionally, various methods from the broader science of operations research and management as well as visualization techniques are employed to precisely model the practical systems. While each paper of this cumulative thesis makes its own contribution theoretically and in a particular decision problem, from a broad viewpoint, the contribution of this work is to create more constructive relationships between decision-making-related disciplines to produce accurate, effective and efficient solutions of multi-criteria problems.