Research Of DEA Alternative-consensus Optimization Model Based On The Discriminating Factor

In consensus decision making, the moderator’s participation improves the speed of reaching consensus and the quality of consensus. The moderator provides compensation value to individual decision makers (DMs) that change their opinions to reach consensus, which faster helps group reach consensus with higher quality.Obviously, the moderator influences the choice of the optimal consensus alternative.To reach consensus, DMs change their opinions and they also expect to gain the compensation values from the moderator. The target of group consensus decision making is achieving the maximum compensation value of all the individual DMs,and the differences in weights among individual DMs are constraints. We construct two types of alternative-consensus optimization models that concern crisp numbers and interval for the DMs’ opinions. By introducing the DEA method into the alternative consensus optimization problems, we construct the alternative-consensus(interval) data envelopment analysis (DEA) model with variable weights: take alternatives as the decision-making units and the optimal alternative is the decision-making unit with the highest output utility, which is also the maximum consensus alternative; the decision-making alternatives are considered as an entire system, whose essence is the DMs’ overall output utility on alternative optimization.Based on the influence of discriminating factor on optimal consensus scheme, we construct the alternative-consensus (interval) optimization model based on the maximum discriminating factor. The alternative-consensus (interval) optimization model based on the maximum discriminating factor obtains the fixed weights of the DMs on different alternatives and overcomes the drawback that the optimal alternative is not unique in the alternative-consensus (interval) DEA model with variable weights. In addition, we build the interval DEA model based on whitening value by introducing the whitening value into the interval DEA, which provides theoretical matting for alternative-consensus interval DEA model with variable weights and the alternative-consensus interval optimization model based on the maximum discriminating factor. Finally, this paper takes the emission reduction problem in the manufacturing industry as an example to illustrate the applicability of the proposed models.