A Classification-Based Consensus Decision-Making Approach with Unknown Class Information for Portfolio Selection Under Hesitant Fuzzy Linguistic Environment

In the portfolio selection process, experts often identify superior alternatives through negotiation and subsequently determine the portfolio structure and ratio. The trust relationships between experts and their divergent opinions on alternative classifications can influence portfolio results, which are rarely considered in existing studies. In this study, we consider the cognitive and affective trust degree of experts and propose a classification-based consensus decision-making (CCDM) approach for portfolio selection. Initially, a cognitive-affective trust network is proposed to determine experts' weights. Subsequently, assuming that the class information, including the number of classifications and that of alternatives in each classification assigned by experts, is unknown, a classification-based consensus-reaching process (CCRP) based on hesitant fuzzy linguistic preference relations (HFLPRs) is proposed for the consensual classifications of alternatives. To obtain the optimal portfolio structure and ratio, a portfolio optimization model that considers realistic constraints is developed. Finally, an application example, comparative analysis, and sensitivity analysis are presented to verify the feasibility and effectiveness of the proposed approach.