Parallelizable PARAFAC decomposition of 3-way tensors

Author

Viet-Dung Nguyen ; Abed-Meraim, Karim ; Linh-Trung Nguyen

Author_Institution

PRISME Lab., Univ. of Orleans, Orleans, France

fYear

2015

fDate

19-24 April 2015

Firstpage

5505

Lastpage

5509

Abstract

This paper introduces a new PARAFAC algorithm for a class of third-order tensors. Particularly, the proposed algorithm is based on subspace estimation and solving a non-symmetrical joint diagonalization problem. To deal with large scale problem, a procedure for overcoming scale and permutation ambiguities is proposed in a parallel computing scheme leading to a significant cost reduction of our method. Performance comparison with some state-of-the-art algorithms produces promising results.

Keywords

cost reduction; signal processing; singular value decomposition; tensors; 3-way tensor; PARAFAC decomposition algorithm; cost reduction; nonsymmetrical joint diagonalization problem; parallel computing scheme; subspace estimation; third-order tensor; Complexity theory; Joints; Loading; Matrix decomposition; Signal processing; Signal processing algorithms; Tensile stress; PARAFAC; Tensor decomposition; non-symmetrical joint diagonalization; parallel and distributed computation; subspace estimation;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on

Conference_Location

South Brisbane, QLD

Type

conf

DOI

10.1109/ICASSP.2015.7179024

Filename

7179024

Link To Document

https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=730864