Adaptive Weighted Graph Fusion Incomplete Multi-View Subspace Clustering

Zhang, Pei; Wang, Siwei; Hu, Jingtao; Cheng, Zhen; Guo, Xifeng; Zhu, En; Cai, Zhiping

doi:10.3390/s20205755

Cited by 19 publications

(7 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…( 7) AGL [34] first constructs complete graphs of observed entries corresponding to each view, then extracts their partition information into a consensus representation. (8) AWGF [38] utilizes feature extraction and incomplete graph fusion in a framework. A sparse regularization is employed to boost clustering performance.…”

Section: Methodsmentioning

confidence: 99%

Self-Representation Subspace Clustering for Incomplete Multi-view Data

Liu

Zhang

et al. 2021

Proceedings of the 29th ACM International Conference on Multimedia

Self Cite

View full text Add to dashboard Cite

Incomplete multi-view clustering is an important research topic in multimedia where partial data entries of one or more views are missing. Current subspace clustering approaches mostly employ matrix factorization on the observed feature matrices to address this issue. Meanwhile, self-representation technique is left unexplored, since it explicitly relies on full data entries to construct the coefficient matrix, which is contradictory to the incomplete data setting. However, it is widely observed that self-representation subspace method enjoys a better clustering performance over the factorization based one. Therefore, we adapt it to incomplete data by jointly performing data imputation and self-representation learning. To the best of our knowledge, this is the first attempt in incomplete multi-view clustering literature. Besides, the proposed method is carefully compared with current advances in experiment with respect to different missing ratios, verifying its effectiveness. CCS CONCEPTS• Computing methodologies → Cluster analysis; Statistical relational learning; Spectral methods; • Theory of computation → Unsupervised learning and clustering; • Mathematics of computing → Nonconvex optimization.

show abstract

Section: Methodsmentioning

confidence: 99%

Self-Representation Subspace Clustering for Incomplete Multi-view Data

Liu

Zhang

et al. 2021

Proceedings of the 29th ACM International Conference on Multimedia

Self Cite

View full text Add to dashboard Cite

show abstract

“…ONKC [14] enhances representability of the optimal kernel and strengthens negotiation between kernel learning and clustering. (3) Graph learning based IMVC [48,49]. IMG [49] extends PVC by exploring rich multi-view global structural information.…”

Section: Incomplete Multi-view Clusteringmentioning

confidence: 99%

Deep Incomplete Multi-view Clustering with Cross-view Partial Sample and Prototype Alignment

Jin¹,

Wang²,

Dong³

et al. 2023

Preprint

View full text Add to dashboard Cite

The success of existing multi-view clustering relies on the assumption of sample integrity across multiple views. However, in real-world scenarios, samples of multi-view are partially available due to data corruption or sensor failure, which leads to incomplete multi-view clustering study (IMVC). Although several attempts have been proposed to address IMVC, they suffer from the following drawbacks: i) Existing methods mainly adopt cross-view contrastive learning forcing the representations of each sample across views to be exactly the same, which might ignore view discrepancy and flexibility in representations; ii) Due to the absence of non-observed samples across multiple views, the obtained prototypes of clusters might be unaligned and biased, leading to incorrect fusion. To address the above issues, we propose a Cross-view Partial Sample and Prototype Alignment Network (CPSPAN) for Deep Incomplete Multi-view Clustering. Firstly, unlike existing contrastive-based methods, we adopt pair-observed data alignment as 'proxy supervised signals' to guide instance-to-instance correspondence construction among views. Then, regarding of the shifted prototypes in IMVC, we further propose a prototype alignment module to achieve incomplete distribution calibration across views. Extensive experimental results showcase the effectiveness of our proposed modules, attaining noteworthy performance improvements when compared to existing IMVC competitors on benchmark datasets.

show abstract

“…Incomplete multi-view clustering algorithms are gradually applied to practical problems, making incomplete multi-view clustering one of the hottest research directions in this area. Traditional IMVC methods can be roughly divided into four categories, i.e., non-negative matrix factorizationbased methods [2,3,4], multi-view subspace clustering [5], kernel learning-based methods [6,7], and graph learning-based methods [8,9]. Among them, the method based on non-negative matrix factorization learns a low-dimensional consistent representation through all view information for clustering.…”

Section: Introductionmentioning

confidence: 99%

Deep multi-view graph clustering with incomplete views

Chang,

Zhu,

Miao

2024

Preprint

View full text Add to dashboard Cite

Deep multi-view graph clustering has made good progress in solving large-scale problems. However, existing deep multi-view graph clustering methods suffer from the following issues: (1) How to combine data processing with multi-view clustering in deep learning, (2) How to learn the local and global information of the graph in deep learning. To this end, a novel method, called deep multi-view graph clustering with the incomplete view (DMVGC-IV), is proposed in this paper, which successfully solves the above two difficulties. Specifically, deep metric learning networks are employed on multiple views to obtain the graph structure. It approximately maintains the semantic distance of data points in the subspace. Then, DMVGC-IV combines global and local structures with a graph-fusion layer. By integrating autoencoder’s reconstruction and multi-view graph learning into a unified framework, our model can jointly optimize the cluster label assignments and embeddings suitable for graph clustering. Experiments use five commonly used multi-view data sets and compare them with five advanced multi-view clustering methods to verify the effectiveness of the proposed method.

show abstract

Adaptive Weighted Graph Fusion Incomplete Multi-View Subspace Clustering

Cited by 19 publications

References 46 publications

Self-Representation Subspace Clustering for Incomplete Multi-view Data

Self-Representation Subspace Clustering for Incomplete Multi-view Data

Deep Incomplete Multi-view Clustering with Cross-view Partial Sample and Prototype Alignment

Deep multi-view graph clustering with incomplete views

Contact Info

Product

Resources

About