Fast Self-Guided Multi-View Subspace Clustering

Multi-view subspace clustering is an important topic in cluster analysis. Its aim is to utilize the complementary information conveyed by multiple views of objects to be clustered. Recently, view-shared anchor learning based multi-view clustering methods have been developed to speed up the learning of common data representation. Although widely applied to large-scale scenarios, most of the existing approaches are still faced with two limitations. First, they do not pay sufficient consideration on the negative impact caused by certain noisy views with unclear clustering structures. Second, many of them only focus on the multi-view consistency, yet are incapable of capturing the cross-view diversity. As a result, the learned complementary features may be inaccurate and adversely affect clustering performance. To solve these two challenging issues, we propose a Fast Self-guided Multi-view Subspace Clustering (FSMSC) algorithm which skillfully integrates the view-shared anchor learning and global-guided-local self-guidance learning into a unified model. Such an integration is inspired by the observation that the view with clean clustering structures will play a more crucial role in grouping the clusters when the features of all views are concatenated. Specifically, we first learn a locally-consistent data representation shared by all views in the local learning module, then we learn a globally-discriminative data representation from multi-view concatenated features in the global learning module. Afterwards, a feature selection matrix constrained by the $\ell _{2,1}$ -norm is designed to construct a guidance from global learning to local learning. In this way, the multi-view consistent and diverse information can be simultaneously utilized and the negative impact caused by noisy views can be overcame to some extent. Extensive experiments on different datasets demonstrate the effectiveness of our proposed fast self-guided learning model, and its promising performance compared to both, the state-of-the-art non-deep and deep multi-view clustering algorithms. The code of this paper is available at https://github.com/chenzhe207/FSMSC.

I. Introduction

Clustering analysis has been an essential research problem in machine learning and pattern recognition, which aims at automatically grouping the data points with similar intrinsic properties into the same cluster by unsupervised learning to represent the data more concisely and to facilitate downstream tasks. However, data are often collected from different sources or sensors in real-world applications. For instance, the same news could be reported by multiple news media; the same semantics could be described by different languages. Although each view is informative and has individual properties, different views are complementary and often deliver the same cluster structures. To better collect complementary information among multiple views, many multi-view clustering algorithms have been proposed in recent years, which can be roughly divided into four categories: the multi-view -means clustering [1], [2], [3], [4], the multi-view spectral clustering [5], [6], [7], the multi-view graph clustering [8], [9], [10], [11], [12], [13], [14], [15], the multi-view subspace clustering [16], [17], [18], [19], [20], [21], [22], [23], [24], and the deep multi-view clustering [25], [26], [27], [28], [29], [30]. Among them, multi-view subspace clustering (MvSC) is very prevailing and has attracted massive attention, due to its excellent data representation capability. The hypothesis behind MvSC is that the consensus representation learned from different views emerges from multiple subspaces associated with different clusters. More specifically, the existing MvSC algorithms can be divided into three different branches, including self-representation learning, matrix factorization and view-shared anchor learning.