Research On Clustering And Aggregating Models Of A Large Scope Of Dual-information

With the decision-making tasks more difficult and exchange of information more convenient,decision-making tends to brainstorming ways, and large group decision-making has become the focusof decision-making problems. In decision-making process, most people focus on how to effectivelyaggregate opinions and preference of decision makers to achieve the best possible agreement, andhave not considered the effect of objective information in the process of aggregation. As a result, thegroup decision-making is difficult to be assured of reliability. Thus this paper proposes a large scopegroup decision-making method based on dual-information.Focusing on a large scope of dual-information given by decision makers, this paper discusses anew three-dimensional grey correlation degree, on which an improved group clustering approach isbased. Then, considered the inconsistency of dual-information, cluster center selecting models areestablished based on grey correlation degrees within and among clusters, which would contributemuch to consistent coordination and aggregation of dual-information after clustering.After the large scope of dual-information clustering, the judgment information of each decisionmaker within the same class may be of unacceptable group consistency. Moreover, due to variousfactors in the process of decision-making, the judgment information given by decision makers may bealso of logical inconsistency, that is, judgment matrices of decision makers will be of unacceptableindividual and group consistency. So the individual and group consistency index is defined. Andbased on the definition, a correction model of consistency is established in order to ensure that thereliability of the large group decision making based on dual-information.Then, after consistent correction of preference information in each class, a new decision makingmethod of dual information within a class is discussed and weighting models are based ondual-information’s linkage. Considering the individual consistency and group consistency of decisionmakers, a weighting model of the decision maker is based on three-dimensional consistency vector.And then this paper establishes a weighting model of the attribute according to the principle ofminimum deviation in the dual information, on which the dual information of each class is effectivelyaggregated based. After aggregation, the interval evaluation value is given by each class.Finally, let the interval evaluation value of great conflict give by each class be pieces of intervalvalued evidence and based on the distance definition of evidence, a conflict estimation model isestablished. Considered the conflict between classes and the numbers of members within each class, adiscount factor is defined to fix the class conflict. At last, an evidence aggregation model is established and the final comprehensive evaluation value is obtained.