Show simple item record

dc.contributor.authorHämäläinen, Joonas
dc.contributor.authorKärkkäinen, Tommi
dc.contributor.authorRossi, Tuomo
dc.date.accessioned2021-01-14T14:02:20Z
dc.date.available2021-01-14T14:02:20Z
dc.date.issued2021
dc.identifier.citationHämäläinen, J., Kärkkäinen, T., & Rossi, T. (2021). Improving Scalable K-Means++. <i>Algorithms</i>, <i>14</i>(1), Article 6. <a href="https://doi.org/10.3390/a14010006" target="_blank">https://doi.org/10.3390/a14010006</a>
dc.identifier.otherCONVID_47636982
dc.identifier.urihttps://jyx.jyu.fi/handle/123456789/73628
dc.description.abstractTwo new initialization methods for K-means clustering are proposed. Both proposals are based on applying a divide-and-conquer approach for the K-means‖ type of an initialization strategy. The second proposal also uses multiple lower-dimensional subspaces produced by the random projection method for the initialization. The proposed methods are scalable and can be run in parallel, which make them suitable for initializing large-scale problems. In the experiments, comparison of the proposed methods to the K-means++ and K-means‖ methods is conducted using an extensive set of reference and synthetic large-scale datasets. Concerning the latter, a novel high-dimensional clustering data generation algorithm is given. The experiments show that the proposed methods compare favorably to the state-of-the-art by improving clustering accuracy and the speed of convergence. We also observe that the currently most popular K-means++ initialization behaves like the random one in the very high-dimensional casesen
dc.format.mimetypeapplication/pdf
dc.languageeng
dc.language.isoeng
dc.publisherMDPI AG
dc.relation.ispartofseriesAlgorithms
dc.rightsCC BY 4.0
dc.subject.otherclustering initialization
dc.subject.otherK-means‖
dc.subject.otherK-means++
dc.subject.otherrandom projection
dc.titleImproving Scalable K-Means++
dc.typearticle
dc.identifier.urnURN:NBN:fi:jyu-202101141104
dc.contributor.laitosInformaatioteknologian tiedekuntafi
dc.contributor.laitosFaculty of Information Technologyen
dc.type.urihttp://purl.org/eprint/type/JournalArticle
dc.description.reviewstatuspeerReviewed
dc.relation.issn1999-4893
dc.relation.numberinseries1
dc.relation.volume14
dc.type.versionpublishedVersion
dc.rights.copyright© 2020 by the authors. Licensee MDPI, Basel, Switzerland
dc.rights.accesslevelopenAccessfi
dc.relation.grantnumber315550
dc.relation.grantnumber311877
dc.subject.ysoalgoritmit
dc.subject.ysotiedonlouhinta
dc.subject.ysoklusterianalyysi
dc.subject.ysoalgoritmiikka
dc.format.contentfulltext
jyx.subject.urihttp://www.yso.fi/onto/yso/p14524
jyx.subject.urihttp://www.yso.fi/onto/yso/p5520
jyx.subject.urihttp://www.yso.fi/onto/yso/p27558
jyx.subject.urihttp://www.yso.fi/onto/yso/p3365
dc.rights.urlhttps://creativecommons.org/licenses/by/4.0/
dc.relation.doi10.3390/a14010006
dc.relation.funderSuomen Akatemiafi
dc.relation.funderSuomen Akatemiafi
dc.relation.funderAcademy of Finlanden
dc.relation.funderAcademy of Finlanden
jyx.fundingprogramAkatemiaohjelma, SAfi
jyx.fundingprogramProfilointi, SAfi
jyx.fundingprogramAcademy Programme, AoFen
jyx.fundingprogramResearch profiles, AoFen
jyx.fundinginformationThe work has been supported by the Academy of Finland from the projects 311877 (Demo) and 315550 (HNP-AI).


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

CC BY 4.0
Except where otherwise noted, this item's license is described as CC BY 4.0