Publications
Detailed Information
S3CMTF: Fast, accurate, and scalable method for incomplete coupled matrix-tensor factorization : (SCMTF)-C-3: Fast, accurate, and scalable method for incomplete coupled matrix-tensor factorization
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Choi, Dongjin | - |
dc.contributor.author | Jang, Jun-Gi | - |
dc.contributor.author | Kang, U. | - |
dc.date.accessioned | 2024-05-27T08:36:44Z | - |
dc.date.available | 2024-05-27T08:36:44Z | - |
dc.date.created | 2020-04-21 | - |
dc.date.issued | 2019-06-28 | - |
dc.identifier.citation | PLoS ONE, Vol.14 No.6, p. e0217316 | - |
dc.identifier.issn | 1932-6203 | - |
dc.identifier.uri | https://hdl.handle.net/10371/203776 | - |
dc.description.abstract | How can we extract hidden relations from a tensor and a matrix data simultaneously in a fast, accurate, and scalable way? Coupled matrix-tensor factorization ( CMTF) is an important tool for this purpose. Designing an accurate and efficient CMTF method has become more crucial as the size and dimension of real-world data are growing explosively. However, existing methods for CMTF suffer from lack of accuracy, slow running time, and limited scalability. In this paper, we propose (SCMTF)-C-3, a fast, accurate, and scalable CMTF method. In contrast to previous methods which do not handle large sparse tensors and are not parallelizable, (SCMTF)-C-3 provides parallel sparse CMTF by carefully deriving gradient update rules. (SCMTF)-C-3 asynchronously updates partial gradients without expensive locking. We show that our method is guaranteed to converge to a quality solution theoretically and empirically. (SCMTF)-C-3 further boosts the performance by carefully storing intermediate computation and reusing them. We theoretically and empirically show that (SCMTF)-C-3 is the fastest, outperforming existing methods. Experimental results show that (SCMTF)-C-3 is up to 930x faster than existing methods while providing the best accuracy. (SCMTF)-C-3 shows linear scalability on the number of data entries and the number of cores. In addition, we apply (SCMTF)-C-3 to Yelp rating tensor data coupled with 3 additional matrices to discover interesting patterns. | - |
dc.language | 영어 | - |
dc.publisher | Public Library of Science | - |
dc.title | S3CMTF: Fast, accurate, and scalable method for incomplete coupled matrix-tensor factorization | - |
dc.title.alternative | (SCMTF)-C-3: Fast, accurate, and scalable method for incomplete coupled matrix-tensor factorization | - |
dc.type | Article | - |
dc.identifier.doi | 10.1371/journal.pone.0217316 | - |
dc.citation.journaltitle | PLoS ONE | - |
dc.identifier.wosid | 000484915000004 | - |
dc.identifier.scopusid | 2-s2.0-85068959158 | - |
dc.citation.number | 6 | - |
dc.citation.startpage | e0217316 | - |
dc.citation.volume | 14 | - |
dc.description.isOpenAccess | Y | - |
dc.contributor.affiliatedAuthor | Kang, U. | - |
dc.type.docType | Article | - |
dc.description.journalClass | 1 | - |
dc.subject.keywordPlus | ALGORITHMS | - |
dc.subject.keywordPlus | DECOMPOSITIONS | - |
dc.subject.keywordPlus | GRADIENT | - |
- Appears in Collections:
- Files in This Item:
- There are no files associated with this item.
Item View & Download Count
Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.