Fuzzy MCDM for Solving Vendor Selection Problems Based on Relation Map

• Main Authors: Jiann-Liang Yang, 楊建樑
• Other Authors: Huan-Neng Chiu
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/32871587752802961540

博士 === 國立臺灣科技大學 === 工業管理系 === 96 === Vendor selection decisions have been long considered one of the most important functions of the purchase department. In the real-world, selecting appropriate vendors should be considered and evaluated in terms of many different criteria resulting in a vast body of data that are often inaccurate or uncertain. Furthermore, the numbers of evaluation criteria in the real complex problems are often too large to determine dependent or independent relationships and there may be no solutions for satisfying all criteria to achieve the aspired level simultaneously. However, most conventional decision models cannot be considered for clarifying the interrelations among the sub-criteria of a criterion by virtue of additivity and independence assumptions. For this reason, this dissertation is to determine three research topics to explore fuzzy multiple criteria decision making (MCDM) for solving vendor selection problem based on relation map. The first topic is to develop a MCDM optimization with compromise ranking methods to select appropriate vendor, in which a fuzzy Delphi method is applied to determine the weights of criteria. The evaluation results indicate the vendor has the trade-off effects while considering a maximum group gain (benefit) and a minimum individual regret (loss) for a company. The second topic is to formulate an integrated fuzzy MCDM technique for solving vendor selection problems. The interrelation between criteria that was ignored by previous researchers is considered in this dissertation. A relationship map to identify the independence or interdependence of the sub-criteria of a criterion is constructed by using interpretive structural modeling (ISM), and then the fuzzy synthetic performance of each common criterion can also be obtained by applying a non-additive fuzzy integral technique. In addition, the results of a practical application show that the proposed method is more suitable than traditional method, especially when the sub-criteria are interdependent in real situations. The third topic is to employ the decision making trial and evaluation laboratory (DEMATEL) method to visualize the structure of complicated causal relationships between criteria of a system and obtain the influence degree of these criteria. Furthermore, the analytical network process (ANP) method is proposed by Saaty to overcome the problems of interdependence and feedback between criteria. The general methods are easy and useful for conducting the problems above. But in ANP procedures, by using average method (equal cluster- weighted) to obtain the weighted supermatrix seem to be irrational because there are different influence degrees among the criteria. Further, by using the concept of ideal point in this dissertation, the results of a practical application could provide some valuable opinions to decision maker on improve each sub-criterion to reduce the gaps between real performance values and aspired/desired values to achieve the best vendor.