Abstract: In this paper, a method considering decision maker's (DM's) psychological behavior is proposed to solve the hybrid stochastic multiple attribute decision making problem. In the method, the attribute values in the form of discrete random variables, grey random variables or linguistic assessment terms are first unified into interval. Then, the aspiration levels proposed by DM with respect to attributes in different phases are regarded as the reference points, and the loss and gain matrix is built by calculating the distance between attribute values and reference points. Further, the gain and loss values of alternatives concerning different attributes are calculated according to cumulative prospect theory. Moreover, the overall prospect values of alternatives are determined by aggregating the values of concerning different attributes in different phases, and a ranking of alternatives can be determined according to the obtained overall prospect values. Finally, a numerical example is used to illustrate the validity and feasibility of the proposed method.

Key words: stochastic multiple attribute decision making, psychological behavior, cumulative prospect theory, alternative ranking