Number of the records: 1
Partitioned Alternating Least Squares Technique for Canonical Polyadic Tensor Decomposition
- 1.
SYSNO ASEP 0460710 Document Type J - Journal Article R&D Document Type Journal Article Subsidiary J Článek ve WOS Title Partitioned Alternating Least Squares Technique for Canonical Polyadic Tensor Decomposition Author(s) Tichavský, Petr (UTIA-B) RID, ORCID
Phan, A. H. (JP)
Cichocki, A. (JP)Number of authors 3 Source Title IEEE Signal Processing Letters. - : Institute of Electrical and Electronics Engineers - ISSN 1070-9908
Roč. 23, č. 7 (2016), s. 993-997Number of pages 5 s. Publication form Print - P Language eng - English Country US - United States Keywords canonical polyadic decomposition ; PARAFAC ; tensor decomposition Subject RIV BB - Applied Statistics, Operational Research R&D Projects GA14-13713S GA ČR - Czech Science Foundation (CSF) Institutional support UTIA-B - RVO:67985556 UT WOS 000379694800005 EID SCOPUS 84978100769 DOI 10.1109/LSP.2016.2577383 Annotation Canonical polyadic decomposition (CPD), also known as parallel factor analysis, is a representation of a given tensor as a sum of rank-one components. Traditional method for accomplishing CPD is the alternating least squares (ALS) algorithm. Convergence of ALS is known to be slow, especially when some factor matrices of the tensor contain nearly collinear columns. We propose a novel variant of this technique, in which the factor matrices are partitioned into blocks, and each iteration jointly updates blocks of different factor matrices. Each partial optimization is quadratic and can be done in closed form. The algorithm alternates between different random partitionings of the matrices. As a result, a faster convergence is achieved. Another improvement can be obtained when the method is combined with the enhanced line search of Rajih et al. Complexity per iteration is between those of the ALS and the Levenberg–Marquardt (damped Gauss–Newton) method. Workplace Institute of Information Theory and Automation Contact Markéta Votavová, votavova@utia.cas.cz, Tel.: 266 052 201. Year of Publishing 2017
Number of the records: 1