Multi-Criteria Decision Making Method Under Uncertainty Based On Composite Linguistic Expressions

Due to the facts that complex and uncertain decision environment and the relevance of the real world decision making problems, decision making method based on linguistic information become a common decision method, because of its intuitionistic interpretability. Therefore, the using of fuzzy linguistic preference information on the decision maker’s depiction of nature more in line with the objective things can reflect the decision-maker’s subjective preference better. In some cases, the expert may hesitate among more than one single linguistic term to express his preference, the linguistic approaches may restrict the elicitation of the linguistic information by using just one linguistic term. Thereby, the composite linguistic expressions are introduced in this thesis, these expressions can express the export’s preference better, and applied to multi-objective decision making and group decision making models. The major work is in the following:(1) Composite linguistic expressions are presented based on the comparative linguistic expressions, provides more elaborated linguistic expressions than single linguistic term, and because these expressions are based on context-free grammar, they are more flexibility and more realistic, providing a new linguistic expression method of multi-criteria decision making under uncertainty.(2) When faced with Multi-attribute decision making problem, with the use of context-free grammars, the composite linguistic expressions are transformed into hesitant fuzzy linguistic term sets(HFLTS), then the fuzzy envelope of HFLTS represented by trapezoidal membership function is calculated by OWA operator, effectively simplifies the computing with words processes. Finally, an illustrative example of its application to a supplier selection problem through the use of fuzzy TOPSIS is presented.(3) The composite linguistic expressions are applied to group decision making model, due to the complexity of objective things and the different of decision makers’ knowledge, experience, and judgment, the decision group is unable to obtain accurate information on the individual preference. When faced with a limited program of multi-objective decision making problems, a collection of different experts’knowledge can better ensure the reasonableness and correctness of the decision making results, and then a new group decision making model based on composite linguistic expressions and corresponding group decision making algorithms are proposed. Firstly with the use of context-free grammars, the linguistic expressions of the preference relations provided by experts are transformed into judgment matrix, and then transform the linguistic expressions of judgment matrix into HFLTS. Finally obtaining a collective preference relation by aggregation in order to rank the alternatives and select the best one.