An almost stochastic dominance based method for stochastic multiple attributes decision making

摘要

This paper proposes an almost stochastic dominance basedmethod for solving the stochastic multiple attribute decision making (SMADM)problems, where the attribute values of alternatives are represented byinterval number with probability distribution (INPD), a random variable withclosed interval support. Firstly, we present the basic operations, setoperations and weighted operators of INPD to accomplish the normalization,weight and aggregation of decision information. Then, based on the conceptof almost stochastic dominance, we propose a new stochastic dominance degree(SDD) to quantify the uncertain relation between INPD. The identification ofideal solution and the measurement of separation are accomplished with SDD.On this basis, an extended TOPSIS method is proposed, includingnormalization method, identification of ideal solution, calculation ofseparation and decision making procedure. Finally, the numerical examplesshow that the new method is effective even under extremely uncertainenvironment, and it can produce more accurate result.