Study Of Clustering Algorithms Based On Continuity

Data mining technology and daily life has become increasingly closer, and people interested in improving about the application of data mining technology.Data mining technology has been widely used in customer relationship management, market trends and forecasts and other areas in the biomedical, financial services, retail, telecommunications and other industries, It provides more valuable information for decision-makers. At the same time, data mining is the need for competition, it provides an important, unknown information and knowledge for policy makers, bringing benefits for decision-makersCluster analysis is a branch of data mining techniques, Not only as an independent data mining tools for data analysis, but also for other data mining methods for data pre-processing, which provides data support for other data mining methods. Clustering process is a collection of physical or abstract objects into object classes similar to the process. Objects have a high similarity in the same class, objects with large differences in different classes. So far, Researchers have proposed a variety of clustering methods for different fields and data types, But they are only appropriate for specific areas. These methods are still many deficiencies in theory and methods.In this paper, Cluster analysis techniques of data mining techniques were analyzed and discussed. First analyzes the clustering method classification and various clustering algorithms for the core technology. And the benifits and weaknesses of the traditional clustering algorithms was explored. Because traditional clustering methods exist in a number of problems, This paper presents a classification method based on the continuity of data under the class definition and constraints. Relative continuum theory has been proposed in discrete state under the principle of continuity. Propose evaluation criteria in discrete state. The use of support and membership to determine clustering results, and through experiments the algorithm was proved.