Data mining for improving the quality of manufacturing: A feature set decomposition approach

Data mining tools can be very beneficial for discovering interesting and useful patterns in complicated manufacturing processes. These patterns can be used, for example, to improve manufacturing quality. However, data accumulated in manufacturing plants have unique characteristics, such as unbalanced distribution of the target attribute, and a small training set relative to the number of input features. Thus, conventional methods are inaccurate in quality improvement cases. Recent research shows, however, that a decomposition tactic may be appropriate here and this paper presents a new feature set decomposition methodology that is capable of dealing with the data characteristics associated with quality improvement. In order to examine the idea, a new algorithm called (Breadth-Oblivious-Wrapper) BOW has been developed. This algorithm performs a breadth first search while using a new F-measure splitting criterion for multiple oblivious trees. The new algorithm was tested on various real-world manufacturing datasets, specifically the food processing industry and integrated circuit fabrication. The obtained results have been compared to other methods, indicating the superiority of the proposed methodology.