论文标题
子空间聚类的内核二维脊回归
Kernel Two-Dimensional Ridge Regression for Subspace Clustering
论文作者
论文摘要
最近对子空间聚类方法进行了广泛的研究。当输入是二维（2D）数据时，现有的子空间聚类方法通常将其转换为向量，从而严重损害了原始数据的固有结构和关系。在本文中，我们为2D数据提出了一种新型的子空间聚类方法。它直接使用2D数据作为输入，从而使表示形式的学习从数据的固有结构和关系中受益。它同时寻求图像投影和表示系数，从而相互增强并导致强大的数据表示。开发了一种有效的算法来解决提出的目标函数，并具有可证明的降低和收敛性。广泛的实验结果验证了新方法的有效性。
Subspace clustering methods have been widely studied recently. When the inputs are 2-dimensional (2D) data, existing subspace clustering methods usually convert them into vectors, which severely damages inherent structures and relationships from original data. In this paper, we propose a novel subspace clustering method for 2D data. It directly uses 2D data as inputs such that the learning of representations benefits from inherent structures and relationships of the data. It simultaneously seeks image projection and representation coefficients such that they mutually enhance each other and lead to powerful data representations. An efficient algorithm is developed to solve the proposed objective function with provable decreasing and convergence property. Extensive experimental results verify the effectiveness of the new method.
