An uncertainty-based model approach using the extended F-Entropy weighted 2D-FAHP method

The increasing data complexity, uncertainty and risk factors associated with digitalisation are making decision-making processes more sensitive and multidimensional. Therefore, multi-criteria decision-making (MCDM) methods stand out for their potential to provide systematic support to decision-makers, particularly in environments characterized by uncertainty. However, the success of these methods depends not only on quantitative data but also on the accurate representation of the judgments and beliefs contained in expert opinions. Decision-makers' subjective assessments often lead to conflicts and uncertainties between criteria, which increases the risk of inconsistency in the weighting process. Within the scope of the thesis, in order to ensure that uncertainties between overlapping criteria are represented more accurately, a correction term sensitive to the number of sub-criteria is integrated into the fuzzy entropy method. The entropy weights obtained with this developed correction factor are included as relative importance weight multipliers in the proposed two-dimensional fuzzy Analytic Hierarchy Process (AHP). In the two-dimensional fuzzy AHP based on fuzzy entropy with correction factors, developed based on the alpha-cutting method, decision matrices are created separately for each alpha value to calculate criterion weights, and these weights are integrated into the system in a way that preserves the two-dimensional fuzzy structure. Thus, the importance ranking between criteria can be obtained more reliably. In the proposed approach, as a result of consistency analyses performed on decision matrices, the alpha-cutoff value at which all matrices or only the matrix based on lower limit values are found to be consistent is determined, and this value is used as a reference in reducing the sub-criteria to a two-dimensional structure. In addition, the developed model is linked to the OCTAVE Allegro framework, enabling the criteria to be analysed together in terms of both their importance and relative risk levels. This holistic approach contributes to the development of more sensitive, consistent and applicable decision support models, particularly in the field of digital information security.