Using Artificial Neural Networks to Determine the Qualification of Suppliers for Automobile Manufactures

• Main Authors: Yi-Ting Su, 蘇羿庭
• Other Authors: Chen-wen Yen
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/5qb298

碩士 === 國立中山大學 === 機械與機電工程學系研究所 === 95 === Many parts used by the automobile manufacturers are provided by outside suppliers. Hence, the chain between the automobile manufacturers and their suppliers has been considered very important for the purchasing department of an automobile factory. Finding qualified suppliers that can meet the demands of the automobile manufacturers is thus an important issue. With the application of neural networks, this thesis develops an approach to help determining the qualification of the suppliers. By using data of the known qualified and unqualified suppliers and by setting a number of features to characterize the capability of the suppliers, neural networks are trained to determine the qualification of the suppliers. In training the neural networks, the features are incrementally removed until optimal classification accuracy is reached. It is hoped that this system can become an effective decision-supporting system in screening the potential suppliers for the automobile manufacturers.