Robust Graph Learning for Multi-view Clustering

Abstract

The multi-view algorithm based on graph learning pays attention to the manifold structure of data and shows the good performance in clustering task. However, multi-view data usually contains noise, which reduces the robustness of multi-view clustering algorithm. In order to solve this problem, we propose a novel multi-view clustering model, namely robust graph learning for multi-view clustering (RGLMC). RGLMC eliminates noise and errors from the original data and employs the adaptive graph, which characterizes the relationship between clusters, as the new input of the algorithm. Our model can be optimized efficiently by utilizing the Augmented Lagrangian Multiplier with Alternating Direction Minimization (ALM-ADM) algorithm. Extensive experimental results on six benchmark datasets verify the superiority of the proposed method in clustering task.