دانلود رایگان مقاله نزدیک شدن به تصمیم گیری با روابط اولویت زبانی بر اساس ثبات افزایشی

Abstract

The linguistic preference relation (LPR) is introduced to efficiently deal with situations in which the decision makers (DMs) provide their preference information by using linguistic labels over paired comparisons of alternatives. However, the lack of consistency in decision making with LPRs can lead to inconsistent conclusions. In this paper, two new decision making methods are developed to improve the additive consistency of LPRs until they are acceptable, and eventually obtain the reliable decision making results. First, the new concepts of order consistency and additive consistency of LPRs are introduced, and followed by a discussion of the characterization about additive consistent LPRs. Then, a consistency index is defined to measure whether an LPR is of acceptable additive consistency. For an unacceptable additive consistent LPR, two automatic iterative algorithms are further proposed to help DMs improve additive consistency level until it is acceptable. In addition, the proposed algorithms can derive the priority weight vector from LPRs and obtain the ranking of the alternatives. Finally, the proposed methods are applied to an emergency operating center (EOC) selection problem. The comparative analysis demonstrates the applicability and effectiveness of the proposed methods.

6. Conclusions In this paper, we first introduce some new concepts, including LPR, order consistent LPR and additive consistent LPR, and the char- acterization about the additive consistency of LPRs is proposed. A consistency index of LPR is defined to measure whether a LPR is of acceptable additive consistency. Moreover, two automatic iterative algorithms are developed to improve LPR with unacceptable additive consistency until the adjusted LPR is acceptably additive consistent. The corresponding automatic iterative algorithms can help the DMs provide the acceptable consistent preferences so as to guarantee the reasonable and identified decision results. In the end, a numerical example is supplied to illustrate the effectiveness and practicality of the developed methods. Comparative analysis are also provided to discuss the performances of our approaches. On the whole, the methodology and algorithm presented in this paper are very important for the application of LPRs in decision making. In terms of future work, we will focus on investigating the multiplicative consistency and consensus reaching models of LPRs on the basis of the results in this paper. Besides, we also intend to apply our methods to the fields of decision making, such as pattern recognition and medical diagnosis, etc.