A Data Mining Framework with Decision Tree and Association Rules and Two Empirical Studies

• Main Authors: Jing-Ching Yang, 楊景晴
• Other Authors: Chen-Fu Chien
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/9v4s87

碩士 === 國立清華大學 === 工業工程與工程管理學系 === 92 === As the progress of IT and computer science, the companies begin to build their own database to store and record the data and to obtain more exact benefits. On the other hand, owing to the complexity, it is more difficult to obtain the information they want that lower the value of the data. Data mining can explore useful information and analyze it from a great deal of data in automative and semi-automative ways. In morden business environment of e-Manufacturing and e-Commerce, the decision maker can make decision based on valuable information and hidden pattern from data mining methods. This research aims to integrate the methodology of Decision Tree and Association Rule and build a architecture of data mining which is suitable for common troubleshooting problem-solving event. Finally, we have the real case studies on distribution feeder faults of Taiwan Power Company and actual engineering data of one semiconducter foundry. The result shows practical viability of data mining approach for troubleshooting.