A dynamic trust consensus model considering individual overconfidence

Abstract

In the consensus reaching process (CRP) under the dynamic trust network, decision-makers’ (DMs’) self-confidence can be regarded as the psychological expression of self-estimation. However, their overconfidence usually has a negative impact on the CRP, even leading to the failure of group decision-making (GDM). To address the issue, we propose a dynamic trust consensus model considering overconfidence under the hesitant intuitionistic fuzzy preference relations (HIFPRs). To determine dynamic trust relationships between multi-attribute DMs, an interactive evaluation-based method is proposed, and thereafter, DMs’ weights are allocated based on the trust degree. In the CRP considering overconfidence, DM’s self-confidence level is defined to detect overconfidence, and feedback mechanisms are constructed for different cases. If overconfident DMs are identified, their trust relationships and self-confidence levels will be adjusted to facilitate the CRP; otherwise, DMs’ HIFPRs will be adjusted. A numerical example, comparative analysis, and discussions are used to illustrate the validity of the proposed model. Comparisons reveal that managing overconfidence by adjusting DMs’ interactive evaluation and self-confidence levels in the dynamic trust network can effectively improve the group consensus degree.