利用菲波那奇数列对栅格尺度进行确定,结合边界确定方式和焦点定位方法获得一种有效的高维数据收缩聚类方法.从基于密度的聚类方法出发,研究密度跨距基础上的良好聚类时变栅格的获取方式,引入菲波那奇数列获取变栅格相应的网格;在实际数据聚类的过程中,采用数据点吸引的方式确定数据质心,通过数据移动和收缩获得最终聚类的焦点和形状.仿真实验的结果表明,本方法可以有效运用于二维和高维数据的聚类分析,在准确获取现有数据内部结构的基础上,具有一定的防噪声作用,可以依据焦点实现数据的有效收缩.%The Febernaci sequence is used to assure the flexible-size grid to get the efficient high dimension shrinking clustering method with the boundary line and foucus locating method. The clustering method based on the density to study the good clustering which was based on the density to get flexible-size grid and febernaci sequence. In the process of clustering method, got the centroid of data by the method of data point to attract to get the focus and shape of movement. The experiment prove that this method can realize shrinking based on the structure of data that prevent noise.
展开▼
