Näytä suppeat kuvailutiedot

dc.contributor.authorWang, Deqing
dc.contributor.authorChang, Zheng
dc.contributor.authorCong, Fengyu
dc.date.accessioned2021-10-07T06:48:51Z
dc.date.available2021-10-07T06:48:51Z
dc.date.issued2021
dc.identifier.citationWang, D., Chang, Z., & Cong, F. (2021). Sparse nonnegative tensor decomposition using proximal algorithm and inexact block coordinate descent scheme. <i>Neural Computing and Applications</i>, <i>33</i>(24), 17369-17387. <a href="https://doi.org/10.1007/s00521-021-06325-8" target="_blank">https://doi.org/10.1007/s00521-021-06325-8</a>
dc.identifier.otherCONVID_101378003
dc.identifier.urihttps://jyx.jyu.fi/handle/123456789/78051
dc.description.abstractNonnegative tensor decomposition is a versatile tool for multiway data analysis, by which the extracted components are nonnegative and usually sparse. Nevertheless, the sparsity is only a side effect and cannot be explicitly controlled without additional regularization. In this paper, we investigated the nonnegative CANDECOMP/PARAFAC (NCP) decomposition with the sparse regularization item using l1-norm (sparse NCP). When high sparsity is imposed, the factor matrices will contain more zero components and will not be of full column rank. Thus, the sparse NCP is prone to rank deficiency, and the algorithms of sparse NCP may not converge. In this paper, we proposed a novel model of sparse NCP with the proximal algorithm. The subproblems in the new model are strongly convex in the block coordinate descent (BCD) framework. Therefore, the new sparse NCP provides a full column rank condition and guarantees to converge to a stationary point. In addition, we proposed an inexact BCD scheme for sparse NCP, where each subproblem is updated multiple times to speed up the computation. In order to prove the effectiveness and efficiency of the sparse NCP with the proximal algorithm, we employed two optimization algorithms to solve the model, including inexact alternating nonnegative quadratic programming and inexact hierarchical alternating least squares. We evaluated the proposed sparse NCP methods by experiments on synthetic, real-world, small-scale, and large-scale tensor data. The experimental results demonstrate that our proposed algorithms can efficiently impose sparsity on factor matrices, extract meaningful sparse components, and outperform state-of-the-art methods.en
dc.format.mimetypeapplication/pdf
dc.language.isoeng
dc.publisherSpringer
dc.relation.ispartofseriesNeural Computing and Applications
dc.rightsCC BY 4.0
dc.subject.othertensor decomposition
dc.subject.othernonnegative CANDECOMP/PARAFAC decomposition
dc.subject.othersparse regularization
dc.subject.otherproximal algorithm
dc.subject.otherinexact block coordinate descent
dc.titleSparse nonnegative tensor decomposition using proximal algorithm and inexact block coordinate descent scheme
dc.typearticle
dc.identifier.urnURN:NBN:fi:jyu-202110075098
dc.contributor.laitosInformaatioteknologian tiedekuntafi
dc.contributor.laitosFaculty of Information Technologyen
dc.contributor.oppiaineSecure Communications Engineering and Signal Processingfi
dc.contributor.oppiaineTekniikkafi
dc.contributor.oppiaineTietotekniikkafi
dc.contributor.oppiaineSecure Communications Engineering and Signal Processingen
dc.contributor.oppiaineEngineeringen
dc.contributor.oppiaineMathematical Information Technologyen
dc.type.urihttp://purl.org/eprint/type/JournalArticle
dc.type.coarhttp://purl.org/coar/resource_type/c_2df8fbb1
dc.description.reviewstatuspeerReviewed
dc.format.pagerange17369-17387
dc.relation.issn0941-0643
dc.relation.numberinseries24
dc.relation.volume33
dc.type.versionpublishedVersion
dc.rights.copyright© The Author(s) 2021
dc.rights.accesslevelopenAccessfi
dc.subject.ysoalgoritmit
dc.subject.ysosignaalinkäsittely
dc.format.contentfulltext
jyx.subject.urihttp://www.yso.fi/onto/yso/p14524
jyx.subject.urihttp://www.yso.fi/onto/yso/p12266
dc.rights.urlhttps://creativecommons.org/licenses/by/4.0/
dc.relation.doi10.1007/s00521-021-06325-8
jyx.fundinginformationOpen access funding provided by University of Jyväskylä (JYU). This work was supported by National Natural Science Foundation of China (Grant No.91748105), National Foundation in China (No. JCKY2019110B009 & 2020-JCJQ-JJ-252), the Fundamental Research Funds for the Central Universities [DUT20LAB303 & DUT20LAB308] in Dalian University of Technology in China, and the scholarship from China Scholarship Council (No. 201600090043).
dc.type.okmA1


Aineistoon kuuluvat tiedostot

Thumbnail

Aineisto kuuluu seuraaviin kokoelmiin

Näytä suppeat kuvailutiedot

CC BY 4.0
Ellei muuten mainita, aineiston lisenssi on CC BY 4.0