Study On Approaches To Multi-attribute Decision Making With Interval Weights

As an important component of the modern decision-making science, the multi-attribute decision-making is widely used in many fields such as engineering design, economic, management and military etc. During the process of decision making, owing to the vagueness and uncertainty of the objective things, large quantity of uncertain data need to be processed. Therefore, it is theoretically and practically significant to make a research on the approaches to the multi-attribute decision making with uncertainty. It's an important content to determine the attribute weights during the multi-attribute decision-making. Duo to the uncertainty and vagueness of the attribute weights, it's much better to adopt interval number and linguistic to express the information of attribute weights for decision maker.The thesis first introduces the basic knowledge of fuzzy sets and interval number. In term of the interval weight vector, it should be considered whether it is normalized interval weight vector or not, then the formula proposes in this paper could be applied to transfer them into normalized interval weight vector. The thesis introduces the approaches to multi-attribute decision-making when attribute weights are certain, the attribute values expresse as exact numerical values and interval form.This paper researches the problem of multi-attribute decision-making when attribute weights are uncertain, the attribute values expresses as exact numerical values The approach is given based on the rule of interval number and the degree of preference of interval number. And it also researches the problem of multi-attribute decision-making when attribute weights are uncertain, the attribute values expresses as exact numerical values. The goal programming model is established to make the decision. Then examples are used to verify their effectiveness and validity. When the information about attribute weights is given with the form of interval-valued intuitionistic preference relation, the notion of multiplicative consistent interval-valued intuitionistic preference relation is introduced first. Secondly, an approach is proposed for obtaining the range of the interval weights based on interval-valued intuitionistic preference relation, also the goal programming models are established. Lastly, two examples are used to verify their effectiveness based on the multiplicative consistent interval-valued intuitionistic preference relation and inconsistent interval-valued intuitionistic preference relation.