The Application Research Of Clustering Analysis In The Six Sigma Project Analysis

The quality management theory of Six Sigma is widely used in China. In the implementation of Six Sigma Project, data analysis is an important part. In order to distinguish the type of data, the data of research object should be classified. When the rule of classification is not given, the method of clustering analysis will be used. Moreover, traditional analysis method of Six Sigma, such as SPC, mostly monitor single index by the control chart. However, when the number of index or variable is large, lots of control charts should be drawn up. The analysis and management of data will be very difficult. So this method of SPC has defect and need to combine with the method of clustering analysis to improve it. Therefore, the research key points of this article are the research of clustering analysis method used in data analysis of Six Sigma, the combination between the Six Sigma analysis method and the clustering analysis method and adding more functions into SMAS (a management and analysis system of Six Sigma project).The theoretical basis of this paper is the quality management theory of Six Sigma and the theory of clustering analysis. This paper introduces the origin and development process of the quality management theory of Six Sigma, DMAIC and common analysis methods of Six Sigma. Then, it introduces the theory of Statistical Process Control, the analysis methods of control chart and the method to judge the process capability. At the same time, it introduces the theory of clustering analysis, the classification of clustering algorithms. The concepts of the distance measure and similarity level are studied in detail. The linkage methods of the hierarchical clustering algorithm are given in the paper.Firstly, this paper studies the clustering analysis algorithm using in data analysis of Six Sigma, including the observation clustering algorithm, variable clustering algorithm and K-means clustering algorithm. The algorithm principle and calculation method is researched in detail.Secondly, an improved analysis method of SPC is studied to solve the defect of SPC. It is based on the clustering analysis. The theory of hypothesis test is used for validation. The experimental analysis is also presented.Finally, this paper designs and implements the clustering analysis system of Six Sigma Project. The architecture of software is based on Web Service. The data analysis tools are developed, which have the functions of the data processing, data computing, clustering analysis, hypothesis test, and the clustering analysis of SPC. And, it is integrated into the system of SMAS. This is a combination of the project management and data analysis. It can expand the analysis functions. At the same time, it is also convenient for data management.