A Study On Data Mining Model To Critical-to-quality Characteristics Identification For Complex Products

With the development of manufacturing, the quality of complex products attractsmore and more people’s attention. As complex products are composed by a largenumber of parts and components, their final quality depends largely on qualitycharacteristics on parts and components’ level. In order to improve the efficiency ofthe quality control process, it is a problem with both theoretical and practicalsignificance that identifying the Critical-to-Quality characteristics among the set ofquality characteristics within intricate relationship between each other.To deal with the problem with high-dimensions, irrelevant and redundant data,the attribute clustering is introduced in this paper to construct the CTQs identifyingmodel based on data mining combining with the method of feature selection.In this paper, related concepts and background of CTQs identifying are reviewedfirstly, then we introduce the fundamental theory of data mining, the tool used in thispaper; in chapter4, the process of constructing model is given, the experiment resultindicates that the model can eliminate the redundancy among quality characteristicseffectively, the following part is a perfection to the model to make it more practical.During the perfection, an index to measure the effectiveness of the result of attributeclustering is proposed. The last chapter is the conclusion of this article, with anoutlook to the research in the future.