A Two-Level Parallel Incremental Tensor Tucker Decomposition Method with Multi-Mode Growth (TPITTD-MG)

With the rapid growth of streaming data, traditional tensor decomposition methods can hardly handle real-time, high-dimensional data of massive amounts in this scenario. In this paper, a two-level parallel incremental tensor Tucker decomposition method with multi-mode growth (TPITTD-MG) is proposed...

Full description

Saved in:
Bibliographic Details
Main Authors: Yajian Zhou, Zongqian Yue, Zhe Chen
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Mathematics
Subjects:
Online Access:https://www.mdpi.com/2227-7390/13/7/1211
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849730187159142400
author Yajian Zhou
Zongqian Yue
Zhe Chen
author_facet Yajian Zhou
Zongqian Yue
Zhe Chen
author_sort Yajian Zhou
collection DOAJ
description With the rapid growth of streaming data, traditional tensor decomposition methods can hardly handle real-time, high-dimensional data of massive amounts in this scenario. In this paper, a two-level parallel incremental tensor Tucker decomposition method with multi-mode growth (TPITTD-MG) is proposed to address the low parallelism issue of the existing Tucker decomposition methods on large-scale, high-dimensional, dynamically growing data. TPITTD-MG involves two mechanisms, i.e., a parallel sub-tensor partitioning mechanism based on the dynamic programming (PSTPA-DP) and a two-level parallel update method for projection matrices and core tensors. The former can count the non-zero elements in a parallel manner and use dynamic programming to partition sub-tensors, which ensures more uniform task allocation. The latter updates the projection matrices or the core tensors by implementing the first level of parallel updates based on the parallel MTTKRP calculation strategy, followed by the second level of parallel updates of different projection matrices or tensors independently based on different classification of sub-tensors. The experimental results show that execution efficiency is improved by nearly 400% and the uniformity of partition results is improved by more than 20% when the data scale reaches an order of magnitude of tens of millions with a parallelism degree of 4, compared with existing algorithms. For third-order tensors, compared with the single-layer update algorithm, execution efficiency is improved by nearly 300%.
format Article
id doaj-art-9887e691197647e299b5d3c4a80b520a
institution DOAJ
issn 2227-7390
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Mathematics
spelling doaj-art-9887e691197647e299b5d3c4a80b520a2025-08-20T03:08:57ZengMDPI AGMathematics2227-73902025-04-01137121110.3390/math13071211A Two-Level Parallel Incremental Tensor Tucker Decomposition Method with Multi-Mode Growth (TPITTD-MG)Yajian Zhou0Zongqian Yue1Zhe Chen2School of Cyberspace Security, Beijing University of Posts and Telecommunications, 1 Nanfeng Rd., Changping District, Beijing 102206, ChinaSchool of Cyberspace Security, Beijing University of Posts and Telecommunications, 1 Nanfeng Rd., Changping District, Beijing 102206, ChinaSchool of Cyberspace Security, Beijing University of Posts and Telecommunications, 1 Nanfeng Rd., Changping District, Beijing 102206, ChinaWith the rapid growth of streaming data, traditional tensor decomposition methods can hardly handle real-time, high-dimensional data of massive amounts in this scenario. In this paper, a two-level parallel incremental tensor Tucker decomposition method with multi-mode growth (TPITTD-MG) is proposed to address the low parallelism issue of the existing Tucker decomposition methods on large-scale, high-dimensional, dynamically growing data. TPITTD-MG involves two mechanisms, i.e., a parallel sub-tensor partitioning mechanism based on the dynamic programming (PSTPA-DP) and a two-level parallel update method for projection matrices and core tensors. The former can count the non-zero elements in a parallel manner and use dynamic programming to partition sub-tensors, which ensures more uniform task allocation. The latter updates the projection matrices or the core tensors by implementing the first level of parallel updates based on the parallel MTTKRP calculation strategy, followed by the second level of parallel updates of different projection matrices or tensors independently based on different classification of sub-tensors. The experimental results show that execution efficiency is improved by nearly 400% and the uniformity of partition results is improved by more than 20% when the data scale reaches an order of magnitude of tens of millions with a parallelism degree of 4, compared with existing algorithms. For third-order tensors, compared with the single-layer update algorithm, execution efficiency is improved by nearly 300%.https://www.mdpi.com/2227-7390/13/7/1211tensorTucker decompositionparallel computingprojection matrixcore tensor
spellingShingle Yajian Zhou
Zongqian Yue
Zhe Chen
A Two-Level Parallel Incremental Tensor Tucker Decomposition Method with Multi-Mode Growth (TPITTD-MG)
Mathematics
tensor
Tucker decomposition
parallel computing
projection matrix
core tensor
title A Two-Level Parallel Incremental Tensor Tucker Decomposition Method with Multi-Mode Growth (TPITTD-MG)
title_full A Two-Level Parallel Incremental Tensor Tucker Decomposition Method with Multi-Mode Growth (TPITTD-MG)
title_fullStr A Two-Level Parallel Incremental Tensor Tucker Decomposition Method with Multi-Mode Growth (TPITTD-MG)
title_full_unstemmed A Two-Level Parallel Incremental Tensor Tucker Decomposition Method with Multi-Mode Growth (TPITTD-MG)
title_short A Two-Level Parallel Incremental Tensor Tucker Decomposition Method with Multi-Mode Growth (TPITTD-MG)
title_sort two level parallel incremental tensor tucker decomposition method with multi mode growth tpittd mg
topic tensor
Tucker decomposition
parallel computing
projection matrix
core tensor
url https://www.mdpi.com/2227-7390/13/7/1211
work_keys_str_mv AT yajianzhou atwolevelparallelincrementaltensortuckerdecompositionmethodwithmultimodegrowthtpittdmg
AT zongqianyue atwolevelparallelincrementaltensortuckerdecompositionmethodwithmultimodegrowthtpittdmg
AT zhechen atwolevelparallelincrementaltensortuckerdecompositionmethodwithmultimodegrowthtpittdmg
AT yajianzhou twolevelparallelincrementaltensortuckerdecompositionmethodwithmultimodegrowthtpittdmg
AT zongqianyue twolevelparallelincrementaltensortuckerdecompositionmethodwithmultimodegrowthtpittdmg
AT zhechen twolevelparallelincrementaltensortuckerdecompositionmethodwithmultimodegrowthtpittdmg