Study on K-means method based on Data-Mining

With the development of computer technology and database management system, more and more data will be accumulated. The problem of the effective usage and selection of these surge data gives birth to a new subject - Data-Mining. Clustering is one of the “Three Data-Mining Technologies”. The K-means algorithm is a simple, practical and efficient clustering algorithm. In this paper, several common clustering algorithm will be simulated combining with real-time data from the power plant boiler. By comparing the quality and execution time of clustering algorithms, the author will select the most suitable algorithm for this kind of data. Finally, the knowledge of clustering will be discovered. And the knowledge is supposed to be applied to industrial applications.