Research On Application Of K-Means Algorithm In User Segmentation

With the rapid development of the society, users demand become more and more diversity.Products that can not meet the demand for personalized will be more a lack of competitiveness. In order to provide such products.first of all, we should dig you user characteristics. This article is based on such a concept, through the statistical method to divide the user community. The K means analysis is one of the most common data mining algorithms. This paper is to study the user segmentation by studying this method..According to the research purpose, this paper will study the user business segments of the domestic telecommunication enterprise. The research topics of this paper is around practice in the implementation of the k-means algorithm to cluster users. Such as variable conversion, exception value processing, K value, etc.The interpretation and application of subdivision results. The whole research is divided into: theoretical analysis stage and empirical analysis stage. The main work of Theory analysis phase is focused on user segmentation and clustering analysis theory; the main work of Empirical analysis phase is discussed the application of K-means algorithm to the specific implementation of the telecommunications users.