Multi-View Bipartite Graph Clustering With Coupled Noisy Feature Filter
- 27 April 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 35 (12) , 12842-12854
Abstract
Unsupervised bipartite graph learning has been a hot topic in multi-view clustering, to tackle the restricted scalability issue of traditional full graph clustering in large-scale applications. However, the existing bipartite graph clustering paradigm pays little attention to the adverse impact of noisy features on learning process. To further facilitate this part of research, apart from simply reweighting features to depress the noisy ones, we take the first step towards analyzing the induced adverse impact via theoretical and experimental investigations. One crucial finding in this article is that the existence of noisy features will incur “anchor shift” phenomenon, which deviates from the potential representations of anchors and then degrades performance. To this end, we propose a coupled noisy feature filter mechanism with automatically finding feature importance to remedy the anchor shift issue in this article. Apart from leveraging features, we theoretically analyze the bounds of proposed feature-adaptive bipartite graph's fuzzy membership. Specifically, distinguishing features’ discrimination will increase the fuzzy membership to achieve soft partitions against the potential inaccurate absolute relationships. With the afore-mentioned merits, our proposed multi-view bipartite graph clustering with coupled noisy feature filter model (MVBGC-NFF) provides novel and interesting insights on the feature level of anchor shift. The effectiveness and efficiency of MVBGC-NFF are demonstrated on synthetic and real-world datasets with improved clustering performance, increasing fuzzy membership, and filtering noisy features. The code is available on https://github.com/liliangnudt/MVBGC-NFF .Keywords
Funding Information
- National Key Research and Development Program of China (2020AAA0107703, 2020AAA0107100)
- National Natural Science Foundation of China (61922088, 61773392, 61976196, 62276271)
This publication has 49 references indexed in Scilit:
- Auto-weighted multi-view clustering via kernelized graph learningPattern Recognition, 2019
- Ultra-Scalable Spectral Clustering and Ensemble ClusteringIEEE Transactions on Knowledge and Data Engineering, 2019
- Incomplete Multiview Spectral Clustering With Adaptive Graph LearningIEEE Transactions on Cybernetics, 2018
- Graph Structure Fusion for Multiview ClusteringIEEE Transactions on Knowledge and Data Engineering, 2018
- Binary Multi-View ClusteringIEEE Transactions on Pattern Analysis and Machine Intelligence, 2018
- Low-Rank Preserving Projection Via Graph Regularized ReconstructionIEEE Transactions on Cybernetics, 2018
- Graph Learning for Multiview ClusteringIEEE Transactions on Cybernetics, 2017
- Self-weighted Multiview Clustering with Multiple Graphs2017
- Large Scale Spectral Clustering Via Landmark-Based Sparse RepresentationIEEE Transactions on Cybernetics, 2014
- LOCUS: learning object classes with unsupervised segmentation2005