Partitioned Hierarchical Alternating Least Squares Algorithm for CP Tensor Decomposition

0472586 - ÚTIA 2018 RIV US eng C - Konferenční příspěvek (zahraniční konf.)
Phan, A. H. - Tichavský, Petr - Cichocki, A.
Partitioned Hierarchical Alternating Least Squares Algorithm for CP Tensor Decomposition.
2017 IEEE International Conference on Acoustics, Speech, and Signal Processing ICASSP 2017. New Orleans: IEEE, 2017, s. 2542-2546. ISBN 978-1-5090-4116-9.
[2017 IEEE International Conference on Acoustics, Speech, and Signal Processing ICASSP 2017. New Orleans (US), 05.03.2017-09.03.2017]
Grant CEP: GA ČR GA17-00902S
Institucionální podpora: RVO:67985556
Klíčová slova: tensor decomposition * canonical polyadic decomposition * PARAFAC * alternating least squares
Obor OECD: Statistics and probability
http://library.utia.cas.cz/separaty/2017/SI/tichavsky-0472586.pdf

Canonical polyadic decomposition (CPD), also known as PARAFAC, is a representation of a given tensor as a sum of rank-one tensors. Traditional method for accomplishing CPD is the alternating least squares (ALS) algorithm. This algorithm is easy to implement with very low computational
complexity per iteration. A disadvantage is that in difficult scenarios, where factor matrices in the decomposition contain nearly collinear columns, the number of iterations needed to achieve convergence might be very large. In this paper, we propose a modification of the algorithm which has similar complexity per iteration as ALS, but in difficult scenarios it needs a significantly lower number of iterations.
Trvalý link: http://hdl.handle.net/11104/0271355