Deep Multi-View Concept Learning

@inproceedings{Xu2018DeepMC,
  title={Deep Multi-View Concept Learning},
  author={Cai Xu and Ziyu Guan and Wei Zhao and Yunfei Niu and Quan Wang and Zhiheng Wang},
  booktitle={IJCAI},
  year={2018}
}
Multi-view data is common in real-world datasets, where different views describe distinct perspectives. To better summarize the consistent and complementary information in multi-view data, researchers have proposed various multi-view representation learning algorithms, typically based on factorization models. However, most previous methods were focused on shallow factorization models which cannot capture the complex hierarchical information. Although a deep multi-view factorization model has… 

Figures and Tables from this paper

Multi-view Clustering via Multiple Auto-Encoder

TLDR
MVC-MAE algorithm adopts Auto-Encoder to capture the non-linear structure information of each view in a layer-wise manner and integrates the representation learning and clustering into a unified step, which jointly optimizes these two steps.

Shared Generative Latent Representation Learning for Multi-view Clustering

TLDR
A novel multi-view clustering method by learning a shared generative latent representation that obeys a mixture of Gaussian distributions that outperforms the state-of-the-art methods under a range of performance criteria.

Co-Learning Non-Negative Correlated and Uncorrelated Features for Multi-View Data

TLDR
A non-negative correlated and uncorrelated feature co-learning (CoUFC) method that identifies view-specific features for each view when learning the common feature across views in the latent semantic subspace and outperforms the state-of-the-art multiview learning methods.

Self-Weighted Multi-View Clustering with Deep Matrix Factorization

TLDR
A novel MVC algorithm based on deep matrix factorization, named Self-Weighted Multi-view Clustering with Deep Matrix Factorization (SMDMF) is presented, which can eliminate interference and reveal semantic information of the multi-view data.

A survey on deepmatrix factorizations

TLDR
This survey paper presents the main models, algorithms, and applications of deep MF through a comprehensive literature review and discusses theoretical questions and perspectives of research as deep MF is likely to become an important paradigm in unsupervised learning in the next few years.

One-step multi-view spectral clustering by learning common and specific nonnegative embeddings

TLDR
This paper proposes a novel one-step multi-view spectral clustering method, in which the spectral embedding and nonnegative embedding are unified into one framework, and can avoid the uncertainty brought by post-processing and obtain optimal clustering result.

Semi-Supervised Multi-Modal Learning with Balanced Spectral Decomposition

TLDR
A Semi-supervised Multimodal Learning Network method (SMLN) which correlates different modalities by capturing the intrinsic structure and discriminative correlation of the multimedia data and will exactly balance all eigenvalues, thus being more competitive to existing methods.

Orthogonal Constrained Meta Heuristic Adaptive Multi-View Clustering Over Multi Labeled Categorical Data Analysis

TLDR
A novel approach i.e. Orthogonal Constrained Meta Heuristic Adaptive Multi-View Clustering (OCMHAMVC) to represent data as a cluster with different categories to provide complexity in computational analysis of data representation.

References

SHOWING 1-10 OF 29 REFERENCES

Multi-View Clustering via Deep Matrix Factorization

TLDR
A deep matrix factorization framework for MVC is presented, where semi-nonnegative matrix factors are adopted to learn the hierarchical semantics of multi-view data in a layer-wise fashion and enforce the non-negative representation of each view in the final layer to be the same.

Multi-View Concept Learning for Data Representation

TLDR
The key idea is to learn a common latent space across different views which captures the semantic relationships between data items through graph embedding regularization on labeled items, and allows each latent factor to be associated with a subset of views via sparseness constraints.

Multi-View Representation Learning: A Survey from Shallow Methods to Deep Methods

TLDR
This survey aims to provide an insightful overview of theoretical basis and state-of-the-art developments in the field of multi-view representation learning and to help researchers find the most appropriate tools for particular applications.

Multi-View Matrix Decomposition: A New Scheme for Exploring Discriminative Information

TLDR
This paper proposes a new multi-view, low-rank, and sparse matrix decomposition scheme to seamlessly integrate diverse yet complementary information stemming from multiple views and enjoys superior performance over the state-of-the-art approaches.

Partially Shared Latent Factor Learning With Multiview Data

TLDR
This paper proposes a novel semisupervised multiview learning algorithm, called partially shared latent factor (PSLF) learning, which jointly exploits both consistent and complementary information among multiple views and introduces a robust sparse regression model to predict the cluster labels of labeled data.

A Deep Matrix Factorization Method for Learning Attribute Representations

TLDR
This work proposes a novel model, Deep Semi-NMF, that is able to learn hidden representations that allow themselves to an interpretation of clustering according to different, unknown attributes of a given dataset, and presents a semi-supervised version of the algorithm, named Deep WSF, that allows the use of (partial) prior information for each of the known Attributes of a dataset.

Learning from Multiple Partially Observed Views - an Application to Multilingual Text Categorization

TLDR
This work addresses the problem of learning classifiers when observations have multiple views, and identifies situations where it is more interesting to use multiple views for learning instead of classical single view learning.

Multi-source Transfer Learning with Multi-view Adaboost

TLDR
This paper proposes a new algorithm named Multi-source Transfer Learning with Multi-view Adaboost (MsTL-MvAdaboost), which not only uses the labeled data from several source tasks to help learn one target task, but also considers how to transfer them in different views synchronously.