Simultaneously Learning Adaptive Neighbors and Clustering Label via Semi-Supervised NMF
Article No.: 15, Pages 1 - 6
Abstract
Graph-based Nonnegative matrix factorization (NMF) methods usually construct an affinity graph in the original high-dimensional space, and then this graph is directly applied to the low-dimensional space, which has effectively improved the clustering performance. However, because of the curse of dimensionality, the graph may not precisely encode the geometric structure of samples in the low-dimensional space. In this paper, we propose a novel semi-supervised NMF method which can simultaneously learn adaptive neighbors and clustering label in the low-dimensional space. Specifically, the high-dimensional data are first projected onto the low-dimensional feature space, where the adaptive neighborhood weights of each feature are exploited to learn a flexible weight graph. Simultaneously, by incorporating the label information, the discriminative feature representations are learned, which assists the weight graph to better capture the potential semantic information. In addition, an efficient alternating minimization algorithm is introduced to solve the proposed model, and extensive experiments are constructed to validate the superiority of the proposed model compared with state-of-the-art methods on five real-word datasets.
References
[1]
Lee D D, Seung H S (1999). Learning the parts of objects by non-negative matrix factorization[J]. Nature, 401(6755), 788--791.
[2]
Liu H, Wu Z, Li X, et al (2012). Constrained Nonnegative Matrix Factorization for Image Representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 34(7), 1299--1311.
[3]
Li Z, Tang J, He X, et al (2018). Robust Structured Nonnegative Matrix Factorization for Image Representation[J]. IEEE Transactions on Neural Networks, 29(5), 1947--1960.
[4]
Lu Y, Yuan C, Zhu W, et al (2018). Structurally Incoherent Low-Rank Nonnegative Matrix Factorization for Image Classification[J]. IEEE Transactions on Image Processing, 27(11), 5248--5260.
[5]
Cai D, He X, Han J, et al (2011). Graph Regularized Nonnegative Matrix Factorization for Data Representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 33(8), 1548--1560.
[6]
Pei X, Wu T, Chen C, et al (2014). Automated Graph Regularized Projective Nonnegative Matrix Factorization for Document Clustering[J]. IEEE Transactions on Systems, Man, and Cybernetics, 44(10), 1821--1831.
[7]
Sun J, Wang Z, Sun F, et al (2018). Sparse dual graph-regularized NMF for image co-clustering[J]. Neurocomputing, 316, 156--165.
[8]
Dong X, Yan Y, Tan M, et al (2019). Late Fusion via Subspace Search With Consistency Preservation[J]. IEEE Transactions on Image Processing, 28(1), 518--528.
[9]
Dong X, Zheng L, Ma F, et al (2018). Few-Example Object Detection with Model Communication[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 41(7), 1641--1654.
[10]
Wu W, Kwong S, Hou J, et al (2019). Simultaneous Dimensionality Reduction and Classification via Dual Embedding Regularized Nonnegative Matrix Factorization[J]. IEEE Transactions on Image Processing, 28 (8), 3836--3847.
[11]
Luo Y, Zhu J, Li M, et al (2018). Smooth Neighbors on Teacher Graphs for Semi-Supervised Learning[J]. computer vision and pattern recognition, 8896--8905.
[12]
Nie F, Wang X, Huang H, et al (2014). Clustering and projected clustering with adaptive neighbors[C]. knowledge discovery and data mining, 977--986.
[13]
Xie D, Zhang X, Gao Q, et al (2019). Multiview Clustering by Joint Latent Representation and Similarity Learning[J]. IEEE Transactions on Systems, Man, and Cybernetics, 1--7.
[14]
Pang Y, Zhou B, Nie F, et al (2019). Simultaneously Learning Neighborship and Projection Matrix for Supervised Dimensionality Reduction[J]. IEEE Transactions on Neural Networks, 30(9),1--15.
[15]
S. P. Boyd and L. Vandenberghe (2004). Convex optimization. Cambridge university press.
[16]
Yang L, Cao X, Jin D, et al (2015). A Unified Semi-Supervised Community Detection Framework Using Latent Space Graph Regularization[J]. IEEE Transactions on Systems, Man, and Cybernetics, 45(11), 2585--2598.
[17]
Wu W, Jia Y, Kwong S, et al (2018). Pairwise Constraint Propagation-Induced Symmetric Nonnegative Matrix Factorization[J]. IEEE Transactions on Neural Networks, 29(12), 6348--6361.
[18]
Kuang D, Choo J, Park H (2015). Nonnegative Matrix Factorization for Interactive Topic Modeling and Document Clustering[J]. Partitional Clustering Algorithms, 215--243.
[19]
Ma Y, Hu X, He T, et al (2016). Hessian regularization based symmetric nonnegative matrix factorization for clustering gene expression and microbiome data[J]. Methods, 111, 80--84.
[20]
Wang H, Yang Y, Li T, et al (2016). Multi-view Clustering via Concept Factorization with Local Manifold Regularization[C]. international conference on data mining, 1245--1250.
Index Terms
- Simultaneously Learning Adaptive Neighbors and Clustering Label via Semi-Supervised NMF
Recommendations
Flexible Adaptive Graph Embedding for Semi-supervised Dimension Reduction
AbstractGraph-based semi-supervised dimension reduction can use the inherent graph structure of samples to propagate label information, and has become a hot research field in machine learning. However, most current methods have strict linear constraints ...
Highlights- A novel semi-supervised adaptive flexible dimension reduction method is proposed.
- The local and global structure of the data is explored by constructing kNN graphs.
- The hard linear constraints on the projection are relaxed by ...
Clustering and projected clustering with adaptive neighbors
KDD '14: Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data miningMany clustering methods partition the data groups based on the input data similarity matrix. Thus, the clustering results highly depend on the data similarity learning. Because the similarity measurement and data clustering are often conducted in two ...
Learning adaptive Grassmann neighbors for image-set analysis
AbstractRepresenting image sets as subspaces on Grassmann manifold and leveraging the Riemannian geometry of this space has proven to be highly effective in various visual recognition tasks. However, the computational cost associated with operating on ...
Highlights- We propose a novel unsupervised DR method to work on non-linear Grassmann manifold.
- We jointly learn the mapping and local structure to alleviate the effect of noise.
- We further measure the distance for Grassmann manifold in terms ...
Comments
Information & Contributors
Information
Published In
October 2019
942 pages
Copyright © 2019 ACM.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected]
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 22 October 2019
Check for updates
Author Tags
Qualifiers
- Research-article
- Research
- Refereed limited
Conference
CSAE 2019
CSAE 2019: The 3rd International Conference on Computer Science and Application Engineering
October 22 - 24, 2019
Sanya, China
Acceptance Rates
Overall Acceptance Rate 368 of 770 submissions, 48%
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 139Total Downloads
- Downloads (Last 12 months)9
- Downloads (Last 6 weeks)2
Reflects downloads up to 28 Feb 2025
Other Metrics
Citations
Cited By
View allView Options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in