Consensus Graph Learning for Incomplete Multi-view Clustering

Zhou, Wei; Wang, Hao; Yang, Yan

doi:10.1007/978-3-030-16148-4_41

Cited by 44 publications

(17 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each learned view-specific graph is constrained to have exactly c (the number of clusters) connected components so that the clustering results can be obtained without requiring any post-clustering. With the help of other views, Zhou et al 42 constructed a complete graph for each view and learned a consensus graph automatically by using each constructed view-specific graph. Zhuge et al 43 connected the processes of spectral embedding and graph completion to recover missing entries of each graph based on multiplications of a common representation matrix and corresponding view-specific representation matrix, and the p-th root is integrated to incorporate losses of multiple views, which weighs the contributions of different views.…”

Section: Related Workmentioning

confidence: 99%

Adaptive partial graph learning and fusion for incomplete multi‐view clustering

et al. 2021

View full text Add to dashboard Cite

Most of existing multi-view clustering methods assume that different feature views of data are fully observed. However, it is common that only portions of data features can be obtained in many practical applications.The presence of incomplete feature views hinders the performance of the conventional multi-view clustering methods to a large extent. Recently proposed incomplete multi-view clustering methods often focus on directly learning a common representation or a consensus affinity similarity graph from available feature views while ignore the valuable information hidden in the missing views. In this study, we present a novel incomplete multi-view clustering method via adaptive partial graph learning and fusion (APGLF), which can capture the local data structure of both within-view and cross-view. Specifically, we use the available data of each view to learn a corresponding view-specific partial graph, in which the within-view local structure can be well preserved. Then we design a cross-view graph fusion term to learn a consensus complete graph for different views, which can take advantage of the complementary information hidden in the viewspecific partial graphs learned from incomplete views.In addition, a rank constraint is imposed on the graph Laplacian matrix of the fused graph to better recover the optimal cluster structure of original data.

show abstract

Section: Related Workmentioning

confidence: 99%

Adaptive partial graph learning and fusion for incomplete multi‐view clustering

et al. 2021

View full text Add to dashboard Cite

show abstract

“…More recently, several graph regularization approaches have been proposed to learn a global or consensus graph from multi-view data for clustering [17], [18] and semi-supervised learning [19]. They employ multi-view features to obtain a unified graph structure.…”

Section: Related Workmentioning

confidence: 99%

“…They employ multi-view features to obtain a unified graph structure. Particularly in [17], [18], the authors propose optimization problems, where single view graph representations are extracted first and then they are fused into a unified graph. In our optimization scheme, we also adopt a graph regularization approach to fit the signal representation model.…”

Section: Related Workmentioning

confidence: 99%

Mask Combination of Multi-layer Graphs for Global Structure Inference

Bayram

Thanou

Vural

et al. 2019

Preprint

View full text Add to dashboard Cite

Structure inference is an important task for network data processing and analysis in data science. In recent years, quite a few approaches have been developed to learn the graph structure underlying a set of observations captured in a data space. Although real-world data is often acquired in settings where relationships are influenced by a priori known rules, this domain knowledge is still not well exploited in structure inference problems. In this paper, we identify the structure of signals defined in a data space whose inner relationships are encoded by multi-layer graphs. We aim at properly exploiting the information originating from each layer to infer the global structure underlying the signals. We thus present a novel method for combining the multiple graphs into a global graph using mask matrices, which are estimated through an optimization problem that accommodates the multi-layer graph information and a signal representation model. The proposed mask combination method also estimates the contribution of each graph layer in the structure of signals. The experiments conducted both on synthetic and real-world data suggest that integrating the multilayer graph representation of the data in the structure inference framework enhances the learning procedure considerably by adapting to the quality and the quantity of the input data.

show abstract

“…Existing traditional incomplete multi-view clustering methods focus on using zero or mean imputation techniques to fill in incomplete information. In this way, Zhou et al [18] proposed to fill missing samples with mean eigenvalues, and then perform clustering on the filled data. However, this padding method may introduce some useless or even noisy information, resulting in poor quality of the construction graph.…”

Section: Introductionmentioning

confidence: 99%

One-Stage Incomplete Multi-view Clustering via Late Fusion

Zhang

Liu

Wang

et al. 2021

Proceedings of the 29th ACM International Conference on Multimedia

View full text Add to dashboard Cite

Incomplete multi-view clustering (IMVC) is an unsupervised approach, among which IMVC via contrastive learning has received attention due to its excellent performance. The previous methods have the following problems: 1) Overreliance on additional projection heads when solving the dimensional collapse problem in which latent features are only valid in lower-dimensional subspaces during clustering. However, many parameters in the projection heads are unnecessary. 2) The recovered view contain inconsistent private information and useless private information will mislead the learning of common semantics due to consistent learning and reconstruction learning on the same feature. To address the above issues, we propose a novel incomplete multi-view contrastive clustering framework. This framework directly optimizes the latent feature subspace, utilizes the learned feature vectors and their subvectors for reconstruction learning and consistency learning, thereby effectively avoiding dimensional collapse without relying on projection heads. Since reconstruction loss and contrastive loss are performed on different features, the adverse effect of useless private information is reduced. For the incomplete data, the missing information is recovered by the cross-view prediction mechanism and the inconsistent information from different views is discarded by the minimum conditional entropy to further avoid the influence of private information. Extensive experimental results of the method on 5 public datasets show that the method achieves state-of-the-art clustering results.

show abstract

Consensus Graph Learning for Incomplete Multi-view Clustering

Cited by 44 publications

References 30 publications

Adaptive partial graph learning and fusion for incomplete multi‐view clustering

Adaptive partial graph learning and fusion for incomplete multi‐view clustering

Mask Combination of Multi-layer Graphs for Global Structure Inference

One-Stage Incomplete Multi-view Clustering via Late Fusion

Contact Info

Product

Resources

About