Application of Data Mining Techniques to Build Product Return Model-A Case Study of Panel Compan

• Main Authors: Chia-Ling Tsou, 鄒佳玲
• Other Authors: 呂奇傑
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/748h8e

碩士 === 健行科技大學 === 工業管理系碩士班 === 104 === In this study, five data mining techniques including discriminant analysis, three decision tree methods (CART, C4.5 and C5.0) and back-propagation neural network (BPN) are used to build product return forecasting model for the panel company. The results showed that the classification accuracy rates of C4.5, C5.0 and BPN are all higher than 70%. Moreover, both of C4.5 and C5.0 techniques can generate the highest classification accuracy rates 76%. Therefore, C4.5 and C5.0 decision tree classification techniques can be used as the reference tools to build product return forecasting model.