基于二进制数计算相似度的高属性维稀疏数据聚类方法

针对一种特定类型高属性维数据———区间变量型高属性维稀疏数据聚类问题,提出高属性维稀疏信息系统,稀疏特征编码,基于二进制数计算相似度概念,给出一种新的基于二进制数计算相似度的高属性维稀疏数据聚类算法,由于计算属性稀疏特征相似度所采用的是二进制数布尔AND运算,因此,相比目前人们所使用的聚类算法,它是一种计算简单、精度高、聚类质量较高的聚类算法。该算法在高属性维稀疏数据挖掘及聚类分析中有着重要的应用。通过数值算例分析表明该聚类方法有效。

High Dimensional Sparse Feature Data Clustering Algorithm Based on Binary Computation

The concepts of high attribute dimensional information s ystem,sparse feature coding and similarity degree base d on binary computation are firstly proposed,and a kind of new high attribute dimensional sparse featur e clustering algorithm.Based on similarity degree of binary Boolean AND computa tion is designed.For similarity degree of sparse feature uses Boolean AND compu tation,compare d with other clustering algorithms ,the algorithm has the chara cteristics of simple computation,high accuracy and efficient clustering,and t he algorithm has been used in data mining.