Research On Multiple Attribute Group Decision Making Approach

The multiple attribute group decision making (MAGDM), which is one of the most important fields on the science of decision making in recent years. This topic is widely used in practice of engineering, economics, marketing, and management. Therefore, systematic study on multiple attribute group decision making method means much for solving practical problems, which is of vital importance in the scientific area of decision making.The multiple attribute group decision making is the crossed research topics that made by the multiple attribute decision making and the group decision making. The problems that MAGDM tries to solve are to aggregate individual judgment to form group judgment, and the decision alternatives are compared, evaluated or ranked by certain decision making technique.The primary research contents and results are listed as follows:(1) For the multiple attribute group decision making problems with the real number attribute values, grey correlation degree is applied to ranking, and compared the results with ranking of the classic TOPSIS (Technique for Order Preference by Similarity to Idea Solution) method. Then the model is built by combining the TOPSIS with grey correlation degree. The result of practical example shows that the model is effective. Moreover, the sensitivity analysis is given. Finally, a new method by combining grey correlation degree with TOPSIS is proposed. The result shows that the method is effective.(2) For the multiple attribute group decision making problems with the interval number attribute values, variable precision rough set is applied to ranking. The essence of rough ranking is changing the interval numbers of decision matrices to real numbers. Then the obtained real number matrices are solved by a new method combined grey correlation degree with TOPSIS (Technique for Order Preference by Similarity to Idea Solution). The result shows that the method is effective. (3) For decision makers' preferences information are characterized by multiplicative and fuzzy preference relations. This paper uses genetic algorithm to solve the CSM (chi-square method) model.