Large-Scale Multi-View Spectral Clustering via Bipartite Graph

By Yeqing Li, Feiping Nie, Heng Huang, Junzhou Huang. AAAI Conference on Artificial Intelligence, 2015

631 citations recorded by the source corpus.

In this paper, we address the problem of large-scale multi-view spectral clustering. In many real-world applications, data can be represented in various heterogeneous features or views. Different views often provide different aspects of information that are complementary to each other. Several previous methods of clustering have demonstrated that better accuracy can be achieved using integrated information of all the views than just using each view individually. One important class of such methods is multi-view spectral clustering, which is based on graph Laplacian. However, existing methods are not applicable to large-scale problem for their high computational complexity. To this end, we propose a novel large-scale multi-view spectral clustering approach based on the bipartite graph. Our method uses local manifold fusion to integrate heterogeneous features. To improve efficiency, we approximate the similarity graphs using bipartite graphs. Furthermore, we show that our method can be easily extended to handle the out-of-sample problem. Extensive experimental results on five benchmark datasets demonstrate the effectiveness and efficiency of the proposed method, where our method runs up to nearly 3000 times faster than the state-of-the-art methods.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/9383f08c697b8aa43782e16c9a57e089911584d8

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Dual Contrastive Prediction for Incomplete Multi-View Representation Learning (2022, methodology): “Caltech101-20 [54]: The dataset consists of 2,386 images of 20 subjects, and we use the HOG and GIST”
Deep Clustering via Joint Convolutional Autoencoder Embedding and Relative Entropy Minimization (2017, background): “Handling large-scale datasets, there are also several studies which select only a subset of data points to accelerate the clustering process [39, 21, 19].”
Multi-Task Multi-View Clustering (2016, methodology): “clustering method in [46] addresses the problem of largescale multi-view spectral clustering.”
Multi-View Clustering and Semi-Supervised Classification with Adaptive Neighbours (2017, methodology): “…(Kumar and III 2011), (b) Co-regularized Spectral Clustering (Co-reg) (Kumar, Rai, and III 2011), (c) Multi-view Spectral Clustering (MVSC) (Li et al. 2015), (d) Multi-Modal Spectral Clustering (MMSC)(Cai et al. 2011), (e) Adaptive Multi-Model Semi-Supervised classification (AMMSS) (Cai et…”
Fast Parameter-Free Multi-View Subspace Clustering With Consensus Anchor Guidance (2021, methodology): “Recently, to improve the efficiency of subspace algorithms, anchor graph strategy has been proposed to select landmarks to represent the entire data items [8], [12], [19]–[23].”
Self-weighted Multiview Clustering with Multiple Graphs (2017, methodology): “In view of this point, some other methods [Li et al., 2015; Xie and Sun, 2013] looked for a new weight learning strategy which has been widely used in multiview learning.”
Multi-view learning overview: Recent progress and new challenges (2017, methodology): “[22] proposed a novel large-scale multi-view spectral clustering approach based on the bipartite graph.”
GMC: Graph-Based Multi-View Clustering (2020, methodology): “For large-scale datasets, the study on large-scale graph-based clustering, such as [17], [57], [58], can be applied to speed up our method.”
Unsupervised Large Graph Embedding (2017, methodology): “To alleviate the problem, inspired by recent progresses on scalable semi-supervised learning (Liu, He, and Chang 2010), large scale spectral clustering (Cai and Chen 2015)(Li et al. 2015), and large scale spectral based dimensionality reduction (Cai 2015), we propose an efficient method, named Unsup”
Incomplete Multiview Spectral Clustering With Adaptive Graph Learning (2020, methodology): “1) Handwritten Digit Database [53]: Following the experimental settings in [53], we conduct the experiments on the multiview handwritten database1 with 2000 samples in total.”
Binary Multi-View Clustering (2019, methodology): “Specifically, following the experimental features used in [11], five different features are selected for Caltech101, i.”
Multi-view clustering: A survey (2018, neutral): “To further uncover inter-view relationships, Formula (8)[82] is rewritten as”
Parameter-Free Auto-Weighted Multiple Graph Learning: A Framework for Multiview Clustering and Semi-Supervised Classification (2016, methodology): “Semi-supervised Classification On different proportions of labeled data (denoted with ⌧ ), we investigate the single graph leaning of Label Propagation (LP) [Zhu et al., 2003] and compare our method with some popular multiple graph leaning methods for semi-supervised classification: Sparse Multiple ”
Multiview Consensus Graph Clustering (2019, background): “Multiview SC via bipartite graph [8] mainly attains high accuracy with low computational complexity but its performance highly relies on the selection of the number of salient points.”
Deep multi-view learning methods: A review (2021, background): “[133] extend it to multi-view scenario as follows 450”
COMPLETER: Incomplete Multi-view Clustering via Contrastive Prediction (2021, background): “In this section, we analyze COMPLETER on the Caltech101-20 dataset from two perspectives, i.e ., parameter sensitivity analysis and ablation studies.”
Graph Learning for Multiview Clustering (2018, neutral): “However, the results of most graphbased MVSC methods are sensitive to the input graph.”