具有稀疏特征的对象—属性子空间边缘重叠区域归属算法

通过分析具有稀疏特征的对象—属性子空间的特征,发现其边缘存在交叉重叠区域现象,为此,提出了基于聚类思想的具有稀疏特征的对象—属性子空间边缘的重叠区域归属算法(OASEDA),该算法能有效解决对象—属性子空间的独立性,算法根据子空间内部紧凑度和子空间之间分离度相对大小确定子空间边缘重叠区域的归属,并基于K-means算法结合权重理论设计了重叠区域归属判断目标函数,最后通过实验证明了该方法的有效性。

Object-attribute subspace with sparse feature edges detection

The overlapped regions among the identified objects-attributes subspaces by the traditional algorithm could influence the independence of these subspaces. In order to solve this defect, this paper developed the objects-attributes subspace edges detection algorithmOASEDAbased on K-means. It designed the objective function of edge detection, algorithm with the information of within-cluster and between-cluster, and optimized the objective function by the weight theory. In the end, experimental results on synthetic datasets demonstrate that the accuracy of the proposed algorithm.