Study On The Methods For Multiple Attribute Group Decision Making Considering Experts' Loss Aversion Behavior

Traditional multiple attribute group decision making (MAGDM) theory and method are based on expected utility theory which assue that experts are completely rational and are able to make optimal decision. However, a lot of facts tell us that experts are always bounded rationality rather than perfect rationality during the process of decision making.The decision result they made is the most suitable for them not the most optimal.So it is necessary to take different psychological behavior of experts into account during the process of MAGDM, as it can impact the decision result.Loss aversion is one of important behaviors in the behavior science.lt refers to people always be more valued "loss" than "profit" when face to the same size probability of selection,namely people are more sensitive to "loss" than "profit".At present,few studies have been conducted to develop scientific and effective decision methods for solving the MAGDM problem considering experts' loss aversion behavior. Therefore, it is of great important theoretic and practical significance to summarize the MAGDM problem considering experts' loss averion behavior and propose targeted methods for the problem.The purpose of this thesis is to study the MAGDM considering experts' loss aversion behavior.With regard to the advantages and disadvantages of existing research, a serious of research works is conducted and contributions are as follows.First, typical MAGDM problems considering experts' loss aversion behavior are extracted-According to the characteristics and the practical backgrounds of MAGDM problem in which experts' loss aversion behavior is considered.typical MAGDM problems considerting experts' loss aversion behavior are extracted in this thesis.Second, provide a measure method of experts' loss aversion degree.For research about this problem, analying the trends of experts' history decision data in the first step. And then according to time of MAGDM compute the weights of loss aversion's degree. Finally, compute composite loss aversion's degeree of experts.Third, propose target methods for integrating multiple attributes evaluation imformation and chosing the optimal alternative.Normalize the decision matrix in the first step.And then based on the TODIM method compute group superiority of alternatives under single attribute.Finally, integrate information of multiple attributes, compute the entire superiority of alternatives and chose the optimal alternative based on this entire superiority.