GROUP DECISION-MAKING PREDICTION MODEL FOR SUPPLIER EVALUATION BASED ON ANFIS

The supplier evaluation process involves criteria such as quality, delivery time, and honesty, typically requiring expert assessment due to its complexity. Consensus-reaching processes (CRPs) are widely studied to support group decisions. CRPs necessitate multiple rounds of expert discussions to achieve consensus, demanding significant resources. The ANFIS technique, which simulates human reasoning through supervised learning and rule construction, has not been applied to automating CRPs in the literature. This paper proposes a model that automates CRPs using ANFIS. Linear regression and paired t-tests indicate that the ANFIS model can predict CRP outcomes without negotiation rounds. Furthermore, the analysis of ANFIS response surfaces demonstrates that the model can graphically represent qualitative factors and identify different decision-maker behaviors during negotiations.