A Co-training Approach for Multi-view Spectral Clustering

By Abhishek Kumar, Hal Daumé Iii. 2011

871 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

A Survey on Multi-view Learning (2013, neutral): “Keywords: Multi-view Learning, Survey, Machine Learning”
Generalized Latent Multi-View Subspace Clustering (2020, background): “Based on spectral clustering, [29], [30] co-regularize the clustering hypothesis of different views to enforce consistence.”
Collaborative Unsupervised Domain Adaptation for Medical Image Diagnosis (2019, background): “unsupervised learning [42], semi-supervised learning [43], [44], weakly-supervised learning [21] and multi-agent learning [45].”
Essential Tensor Learning for Multi-View Spectral Clustering (2018, background): “For example, [18] proposes a co-training approach to search for the clusterings that agree across the views.”
Graph Learning for Multiview Clustering (2018, unknown): “[18], [19] proposed unsupervised multiview clustering methods based on co-training learning, such as cotraining multiview spectral clustering (SC) and co-regularized multiview SC (CRSC).”
Co-regularized Alignment for Unsupervised Domain Adaptation (2018, methodology): “The related ideas of co-training [3] and co-regularization [37, 36] have been successfully used for semi-supervised learning as well as unsupervised learning [21, 20].”
Multiple Incomplete Views Clustering via Weighted Nonnegative Matrix Factorization with L2, 1 Regularization (2015, background): “[23,24] are among the first works proposed to solve the multi-view clustering problem via spectral projection.”
Kernel-Based Weighted Multi-view Clustering (2012, methodology): “In [20] the spectral embedding from one view is used to bootstrap the clustering of the other view, by modifying its similarity matrix in a co-training [3] like fashion.”
Self-weighted Multiview Clustering with Multiple Graphs (2017, methodology): “Inspired by co-training [Blum and Mitchell, 1998], which is widely used in multiview semisupervised learning, [Kumar and Daumé, 2011] searched for the clusters that agree across all the views.”
An Overview of Deep Semi-Supervised Learning (2020, unknown): “In such a setting, MVL aims to learn a distinct prediction function f θ i to model a given view v i ( x ) of a data point x , and jointly optimize all the functions to improve the generalization performance.”
Unifying Virtual and Physical Worlds (2017, methodology): “We treat the scaled community assignment matrices G v and G p as the new user representations [Kumar et al. 2011].”
Multi-Task Multi-View Clustering (2016, methodology): “The multi-view clustering method in [43] applies the idea of co-training, which uses the spectral embedding from one view to constrain the similarity graph used for the other view.”
Robust Multi-View Spectral Clustering via Low-Rank and Sparse Decomposition (2014, methodology): “…quantities of data in multiple representations from diverse sources, multi-view clustering algorithms have attracted considerable attention (Chaudhuri et al. 2009; Bickel and Scheffer 2004; Kumar and Daumé 2011; Kumar, Rai, and Daumé 2011; Zhou and Burges 2007; Greene and Cunningham 2009).”
Partial Multi-View Clustering (2014, background): “Kumar and Daume(2011) propose a cotraining approach where the similarity matrix of one view is constrained by the spectral embedding of the other view.”
A survey of multi-view machine learning (2013, background): “Kumar and Daumé [23] presented a co-training approach for multiview spectral clustering, where the clusterings of different views are bootstrapped using information from one another.”
Multi-View Clustering via Joint Nonnegative Matrix Factorization (2013, methodology): “[15] approaches the problem by adopting an co-training framework [5] such that the similarity matrix in one view is affected by the similarity estimated based on the eigenvectors of Laplacian matrix in the other view.”
Latent Multi-view Subspace Clustering (2017, methodology): “Under the spectral clustering framework, the methods [16, 15] coregularize the hypothesis to be consistent across these different views.”