Reliable Multi-View Clustering

Abstract

With the advent of multi-view data, multi-view learning (MVL) has become an important research direction in machine learning. It is usually expected that multi-view algorithms can obtain better performance than that of merely using a single view. However, previous researches have pointed out that sometimes the utilization of multiple views may even deteriorate the performance. This will be a stumbling block for the practical use of MVL in real applications, especially for tasks requiring high dependability. Thus, it is eager to design reliable multi-view approaches, such that their performance is never degenerated by exploiting multiple views.This issue is vital but rarely studied. In this paper, we focus on clustering and propose the Reliable Multi-View Clustering (RMVC) method. Based on several candidate multi-view clusterings, RMVC maximizes the worst-case performance gain against the best single view clustering, which is equivalently expressed as no label information available. Specifically, employing the squared χ2 distance for clustering comparison makes the formulation of RMVC easy to solve, and an efficient strategy is proposed for optimization. Theoretically, it can be proved that the performance of RMVC will never be significantly decreased under some assumption. Experimental results on a number of data sets demonstrate that the proposed method can effectively improve the reliability of multi-view clustering.

One-step graph-based incomplete multi-view clustering

“…Many clustering methods have been proposed to solve the multi-view clustering task, which can be roughly divided into six categories according to their strategies: multi-kernel clustering [37], [42], co-training style clustering [33], multi-view subspace clustering [12], [5], multi-view spectral clustering [21], [62], deep multi-view clustering, [50], graph-based multi-view clustering [31], [29], [41], [54], [20], which is most related to our works.…”

“…For graph-based multi-view clustering, Nie et al [29] minimized the reconstruction error by learning an optimal weight for each graph to build a unified graph. Tao et al [41] learned a fusion graph by using the information of multi graphs from both view-level and sample-view. Zhan et al [54] learned the cluster indicating matrices to merge the information from singleview graphs, and then minimized disagreement between each individual view and the global view to obtain a consensus graph.…”

Multi-View Adjacency-Constrained Nearest Neighbor Clustering (Student Abstract)

“…Neighbor Clustering (ANNC); and the internal evaluation Index based on Rawls' Max-Min criterion (Kameda et al, 2016) (MMI). Given extracting information from each single view is also important (Tao et al, 2018), FDEW aims to learn a fusion distance matrix set, which not only uses complementary information among multiple views, but exploits the information from each single view. ANNC obeys an intuitive rule that one cluster and its nearest neighbor with higher mass (size) should be grouped into the same cluster in the clustering process (Yang and Lin, 2020).…”