Dynamic clustering method on the basis of sparse feature
暂无分享,去创建一个
The concepts of high attribute dimensional information system are firstly proposed, and a new dynamic clustering method on the basis of sparse feature difference degree is presented. The new method is flexible and sparse feature attribute-oriented. By virtue of sparse feature difference degree, it can dynamically and effectively classify attributes. Moreover, it is fault-tolerant. The algorithm will have important application in high attribute dimensional data mining.