Method for uncertain multiple attribute decision making based on estimation of attribute value functions

With regard to the uncertain multiple attribute decision making problems in which information on attribute weights is unknown completely and attribute values are in real numbers, an analysis method is proposed based on the estimation of attribute value functions. First, the multiple attribute utility functions and value functions are introduced. Then, in order to calculate the values of value functions for each attribute, an approximate curve is constructed to obtain the form of value functions. Further, the overall values of each alternative are obtained by adding the values of value functions for each attribute in this alternative and thus the alternative ranking results can be determined. Finally, an example is used to illustrate the proposed method.