Robust active representation via ℓ<inf>2,p</inf>-norm constraints[Formula presented]

Zhao, J; Yi, S; Liang, Y; Liu, W; Cao, X

Robust active representation via ℓ<inf>2,p</inf>-norm constraints[Formula presented]

Zhao, J Yi, S Liang, Y Liu, W Cao, X

Permalink

Publisher:: Elsevier
Publication Type:: Journal Article
Citation:: Knowledge-Based Systems, 2022, 235, pp. 107639
Issue Date:: 2022-01-10

Closed Access

	Filename	Description	Size
	1-s2.0-S0950705121009011-main.pdf	Published version	1.24 MB		View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Zhao, J
dc.contributor.author	Yi, S
dc.contributor.author	Liang, Y
dc.contributor.author	Liu, W
dc.contributor.author	Cao, X
dc.date.accessioned	2023-07-08T21:16:12Z
dc.date.available	2023-07-08T21:16:12Z
dc.date.issued	2022-01-10
dc.identifier.citation	Knowledge-Based Systems, 2022, 235, pp. 107639
dc.identifier.issn	0950-7051
dc.identifier.uri	http://hdl.handle.net/10453/171353
dc.description.abstract	Active learning maximizes the performance of the current learning model by soliciting benefits from unlabeled data. In its early stage with insufficient labels, finding representations that maintain a consistent hypothesis with the entire unlabeled pool is an intelligent paradigm. From a matrix perspective, this paradigm can be transformed into a sparse representation of the input matrix by the ℓ2,1-norm constraint. However, the ℓ2,1-norm constraint is used in the loss function, which may lead to the mean accumulation problem, resulting in sub-optimal mean-centering and low robustness to outliers. In this paper, to solve the aforementioned problems, we generalize the ℓ2,1-norm into the ℓ2,p-norm constraint, where 0<p≤1. Specifically, when p=1, the ℓ2,1 constraint derives the optimal mean-centering, avoiding biased feature transformations, and when 0<p<1, ℓ2,p regulates the sparsity of the loss function and regularization terms in the matrix representation, yielding robust active representations. Experiment results on benchmark databases demonstrate that our method outperforms other state-of-the-art active learning methods, particularly for outlier perturbations.
dc.language	en
dc.publisher	Elsevier
dc.relation.ispartof	Knowledge-Based Systems
dc.relation.isbasedon	10.1016/j.knosys.2021.107639
dc.rights	info:eu-repo/semantics/closedAccess
dc.subject	08 Information and Computing Sciences, 15 Commerce, Management, Tourism and Services, 17 Psychology and Cognitive Sciences
dc.subject.classification	Artificial Intelligence & Image Processing
dc.subject.classification	4602 Artificial intelligence
dc.subject.classification	4605 Data management and data science
dc.subject.classification	4611 Machine learning
dc.title	Robust active representation via ℓ<inf>2,p</inf>-norm constraints[Formula presented]
dc.type	Journal Article
utslib.citation.volume	235
utslib.for	08 Information and Computing Sciences
utslib.for	15 Commerce, Management, Tourism and Services
utslib.for	17 Psychology and Cognitive Sciences
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology/School of Computer Science
utslib.copyright.status	closed_access	*
dc.date.updated	2023-07-08T21:16:11Z
pubs.publication-status	Published
pubs.volume	235

Abstract:

Active learning maximizes the performance of the current learning model by soliciting benefits from unlabeled data. In its early stage with insufficient labels, finding representations that maintain a consistent hypothesis with the entire unlabeled pool is an intelligent paradigm. From a matrix perspective, this paradigm can be transformed into a sparse representation of the input matrix by the ℓ2,1-norm constraint. However, the ℓ2,1-norm constraint is used in the loss function, which may lead to the mean accumulation problem, resulting in sub-optimal mean-centering and low robustness to outliers. In this paper, to solve the aforementioned problems, we generalize the ℓ2,1-norm into the ℓ2,p-norm constraint, where 0

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/171353