Dominance is an underlying concept in decision making that is used to develop a method to obtain the derived weight. The eigenvector method (EM) [3] is favored because it captures dominance at the level of toler...Dominance is an underlying concept in decision making that is used to develop a method to obtain the derived weight. The eigenvector method (EM) [3] is favored because it captures dominance at the level of tolerated inconsistency, the least included angles method (LAM) [1] minimizes error without an explict attempt to capture dominance, but with simplicity and practicality. This paper puts forward a new priority method——the least included angles method with mean cumulative dominance (DLAM) combining the good characteristics of EM and LAM. Compared with the EM, the LAM, the GMDM and the AMDM, the DLAM is a simpler, practical and more rational method in calculating the weight vectors of judgement matrices. The results of the numerical example also show that the DLAM and the EM always derive the same rankings, the other methods such as the LAM, the logarithmic least squares method (LLSM), the GMDM and the AMDM are possible to obtain the rankings, which are different from those derived by the DLAM and the EM.展开更多
基金Research supported by National Science Foundation of China
文摘Dominance is an underlying concept in decision making that is used to develop a method to obtain the derived weight. The eigenvector method (EM) [3] is favored because it captures dominance at the level of tolerated inconsistency, the least included angles method (LAM) [1] minimizes error without an explict attempt to capture dominance, but with simplicity and practicality. This paper puts forward a new priority method——the least included angles method with mean cumulative dominance (DLAM) combining the good characteristics of EM and LAM. Compared with the EM, the LAM, the GMDM and the AMDM, the DLAM is a simpler, practical and more rational method in calculating the weight vectors of judgement matrices. The results of the numerical example also show that the DLAM and the EM always derive the same rankings, the other methods such as the LAM, the logarithmic least squares method (LLSM), the GMDM and the AMDM are possible to obtain the rankings, which are different from those derived by the DLAM and the EM.
