Three-way Behavioral Decision-making Based On Dominance Relations For Incomplete Information Systems

Typical solutions for dealing with incomplete information systems have been extensively studied,but most of these approaches change the original data information and do not guarantee the integrity of the original information system.Meanwhile,the psychological behavior of decision makers often has an impact on the decision outcome.In this paper,we combine three-way decision with behavioral decision making and propose two decision making approaches based on probability dominance relation in incomplete information systems,which can effectively provide intelligent decision support for solving practical decision making problems in incomplete information systems.The specific research includes the following two aspects:(1)A three-way decision approach with probability dominance relation based on prospect theory is constructed for incomplete information systems.First,the probability dominance relation of incomplete information system is defined to deal with the binary relation between evaluation values,and the calculation approach of relative value functions in the behavioral decision process is given;in addition,the conditional probability with decision attributes in incomplete decision information system is proposed,and the objective weight function is further obtained by using the prospect theory.On this basis,a three-way decision approach with probability dominance relation based on prospect theory is proposed for the incomplete information system.Finally,the stability,effectiveness and superiority of the proposed approach are verified through corresponding experimental studies and comparative analysis of two actual medical cases.(2)An approach of three-way decision with a priori probability tolerance dominance relation based on regret theory in incomplete information systems is established.First,the a priori probability tolerance dominance relation of incomplete information system is defined to deal with the binary relation between evaluation values,and the approach of calculating objective weights is given;in addition,the decision attribute values of each alternative in incomplete decision information system can be calculated,and the integrated utility perception value is obtained by combining with regret theory.Further,a three-way decision approach based on regret theory with a priori probability tolerance dominance relation is proposed for incomplete information systems.Finally,the stability,effectiveness and superiority of the proposed approach are verified through corresponding experimental studies and comparative analysis of two real cases.