Building consensus in multi-attribute group decision making under a prospect theory-driven feedback adjustment mechanism

This paper reports a prospect theory-based consensus model in multi-attribute group decision making. First, the attribute expectations are regarded as the reference points of decision makers in prospect theory. The attribute values are transformed into the prospect values based on the value function. Then a new consensus measure is proposed to quantify the consensus level. A feedback adjustment mechanism is constructed to improve the consensus level under prospect theory. The convergence of the proposed algorithm is proved. It is found that the proposed mechanism can lessen the work burden of DMs as compared with the existing methods. Finally, the weights of attributes are computed by constructing an optimization model that maximizes the sum of prospect values. Case study and comparative analysis are offered to illustrate the novelty and advantage of the proposed model. The obtained results reveal that an increasing loss aversion enhances the difficulty of reaching consensus.