Zobrazit v EDS

Plug-and-play low-rank tensor completion and reconstruction algorithms with improved applicability of tensor decompositions

Uloženo v:
Podrobná bibliografie
Název: Plug-and-play low-rank tensor completion and reconstruction algorithms with improved applicability of tensor decompositions
Autoři: Manabu Mukai, Hidekata Hontani, Tatsuya Yokota
Zdroj: Frontiers in Applied Mathematics and Statistics. 11
Informace o vydavateli: Frontiers Media SA, 2025.
Rok vydání: 2025
Popis: In this paper, we propose a new unified optimization algorithm for general tensor completion and reconstruction problems, which is formulated as an inverse problem for low-rank tensors in general linear observation models. The proposed algorithm supports at least three basic loss functions (ℓ2 loss, ℓ1 loss, and generalized KL divergence) and various TD models (CP, Tucker, TT, TR decompositions, non-negative matrix/tensor factorizations, and other constrained TD models). We derive the optimization algorithm based on a hierarchical combination of the alternating direction method of multipliers (ADMM) and majorization-minimization (MM). We show that the proposed algorithm can solve a wide range of applications and can be easily extended to any established TD model in a plug-and-play manner.
Druh dokumentu: Article
ISSN: 2297-4687
DOI: 10.3389/fams.2025.1594873
Rights: CC BY
Přístupové číslo: edsair.doi...........d9b8ec68f3d893efc0f8f383c6166a7a
Databáze: OpenAIRE
Popis
Abstrakt:In this paper, we propose a new unified optimization algorithm for general tensor completion and reconstruction problems, which is formulated as an inverse problem for low-rank tensors in general linear observation models. The proposed algorithm supports at least three basic loss functions (ℓ2 loss, ℓ1 loss, and generalized KL divergence) and various TD models (CP, Tucker, TT, TR decompositions, non-negative matrix/tensor factorizations, and other constrained TD models). We derive the optimization algorithm based on a hierarchical combination of the alternating direction method of multipliers (ADMM) and majorization-minimization (MM). We show that the proposed algorithm can solve a wide range of applications and can be easily extended to any established TD model in a plug-and-play manner.
ISSN:22974687
DOI:10.3389/fams.2025.1594873