With the rapid development of data collection and storage technology recently, lots of application fields accumulated large amount of data. Data mining technology, combining with database, statistics and machine learning methods, is utilized to find the valuable knowledge and rules from these data. Clustering analysis technology is the classical part in data mining and is an important tool in many research areas.Based on survey of relative papers and further study of clustering algorithm, the following work has been done in clustering improvement and application.1) An automated clustering algorithm is proposed to avoid the disadvantages of classical clustering algorithm, such as shape dependence, reasonably selection of class parameters and sensitivity to noise. The proposed algorithm makes use of data reduction, improved K-means... |