File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1016/j.sigpro.2023.108966
- Scopus: eid_2-s2.0-85148041698
- WOS: WOS:000940674600001
- Find via
Supplementary
- Citations:
- Appears in Collections:
Article: Accelerating probabilistic tensor canonical polyadic decomposition with nonnegative factors: An inexact BCD Approach
| Title | Accelerating probabilistic tensor canonical polyadic decomposition with nonnegative factors: An inexact BCD Approach |
|---|---|
| Authors | |
| Keywords | Automatic tensor rank determination Nonnegative factors Tensor decomposition |
| Issue Date | 1-Jun-2023 |
| Publisher | Elsevier |
| Citation | Signal Processing, 2023, v. 207 How to Cite? |
| Abstract | Recently, Bayesian modeling and variational inference (VI) were leveraged to enable the nonnegative factor matrix learning with automatic rank determination in tensor canonical polyadic decomposition (CPD), which has found various applications in big data analytics. However, since VI inherently performs block coordinate descent (BCD) steps over the functional space, it generally does not allow integration with modern large-scale optimization methods, making the scalability a critical issue. In this paper, it is revealed that the expectations of the variables updated by the VI algorithm is equivalent to the block minimization steps of a deterministic optimization problem. This equivalence further enables the adoption of inexact BCD method for devising a fast nonnegative factor matrix learning algorithm with automatic tensor rank determination. Numerical results using synthetic data and real-world applications show that the performance of the proposed algorithm is comparable with that of the VI-based algorithm, but with computation times reduced significantly. |
| Persistent Identifier | http://hdl.handle.net/10722/339298 |
| ISSN | 2023 Impact Factor: 3.4 2023 SCImago Journal Rankings: 1.065 |
| ISI Accession Number ID |
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.author | Chen, Zhongtao | - |
| dc.contributor.author | Cheng, Lei | - |
| dc.contributor.author | Wu, Yik-Chung | - |
| dc.date.accessioned | 2024-03-11T10:35:30Z | - |
| dc.date.available | 2024-03-11T10:35:30Z | - |
| dc.date.issued | 2023-06-01 | - |
| dc.identifier.citation | Signal Processing, 2023, v. 207 | - |
| dc.identifier.issn | 0165-1684 | - |
| dc.identifier.uri | http://hdl.handle.net/10722/339298 | - |
| dc.description.abstract | <p>Recently, <a href="https://www.sciencedirect.com/topics/computer-science/bayesian-modeling" title="Learn more about Bayesian modeling from ScienceDirect's AI-generated Topic Pages">Bayesian modeling</a> and variational inference (VI) were leveraged to enable the nonnegative <a href="https://www.sciencedirect.com/topics/engineering/matrix-factor" title="Learn more about factor matrix from ScienceDirect's AI-generated Topic Pages">factor matrix</a> learning with automatic rank determination in tensor canonical polyadic decomposition (CPD), which has found various applications in big data analytics. However, since VI inherently performs block coordinate descent (BCD) steps over the functional space, it generally does not allow integration with modern large-scale optimization methods, making the scalability a critical issue. In this paper, it is revealed that the expectations of the variables updated by the VI algorithm is equivalent to the block minimization steps of a deterministic <a href="https://www.sciencedirect.com/topics/computer-science/optimization-problem" title="Learn more about optimization problem from ScienceDirect's AI-generated Topic Pages">optimization problem</a>. This equivalence further enables the adoption of inexact BCD method for devising a fast nonnegative <a href="https://www.sciencedirect.com/topics/engineering/matrix-factor" title="Learn more about factor matrix from ScienceDirect's AI-generated Topic Pages">factor matrix</a> learning algorithm with automatic <a href="https://www.sciencedirect.com/topics/engineering/rank-tensor" title="Learn more about tensor rank from ScienceDirect's AI-generated Topic Pages">tensor rank</a> determination. Numerical results using synthetic data and real-world applications show that the performance of the proposed algorithm is comparable with that of the VI-based algorithm, but with computation times reduced significantly.<br></p> | - |
| dc.language | eng | - |
| dc.publisher | Elsevier | - |
| dc.relation.ispartof | Signal Processing | - |
| dc.subject | Automatic tensor rank determination | - |
| dc.subject | Nonnegative factors | - |
| dc.subject | Tensor decomposition | - |
| dc.title | Accelerating probabilistic tensor canonical polyadic decomposition with nonnegative factors: An inexact BCD Approach | - |
| dc.type | Article | - |
| dc.identifier.doi | 10.1016/j.sigpro.2023.108966 | - |
| dc.identifier.scopus | eid_2-s2.0-85148041698 | - |
| dc.identifier.volume | 207 | - |
| dc.identifier.eissn | 1872-7557 | - |
| dc.identifier.isi | WOS:000940674600001 | - |
| dc.identifier.issnl | 0165-1684 | - |