A Methodological Approach to Aggregation of Expert and Statistically Determined Data in Determining Weighting Factor Values in Decision-Making Problems

A methodology for synthesizing expert and statistically determined weights for constructing integrated indicators is proposed. The approachis based on the adaptive combination of subjective expert assessments and objective entropy weights using a confidence coefficient calculated based on the consistency of the expert group, the variability of assessments, and the sufficiency of the number of experts. A logistic function is used to ensure a «soft» redistribution of influence between weighting sources. The final synthetic weight vectoris formed through a linear combination of weights defined using two methodological bases. The proposed approachensures a balance between the substantive significance of criteria and their statistical informativeness, increases the robustness of the integrated assessment, and allows the method to automatically adapt to the quality of expert information. The method can be applied in decision support systems and in analyzing complex multi-criteria objects.