Improving Quality of Products in Hard Drive Manufacturing by Decision Tree Technique
Anotai Siltepavet, Sukree Sinthupinyo, Prabhas Chongstitvatana · 2012
Hard drives manufacturing is a complex process. The quality of products is determined by a large set of parameters. However, there are defects in products that need to be removed. We study a systematic approach to find suitable parameters which can reduce the number of defected products. Our approach exploits decision tree learning and a set of algorithms to adjust decision parameters obtained from the learned decision tree. Moreover, because we cannot test the result in the real environment, we propose a trustable testing method which can predict the improvement obtained from the parameter adjustment system. The results from the experiments show that the quality of products in the dataset can be improved as much as 12%. Which is significant in hard drive manufacturing.