Fast Alternating LS Algorithms for High Order CANDECOMP/PARAFAC Tensor Factorizations

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F67985556%3A_____%2F13%3A00396774" target="_blank" >RIV/67985556:_____/13:00396774 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1109/TSP.2013.2269903" target="_blank" >http://dx.doi.org/10.1109/TSP.2013.2269903</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/TSP.2013.2269903" target="_blank" >10.1109/TSP.2013.2269903</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Fast Alternating LS Algorithms for High Order CANDECOMP/PARAFAC Tensor Factorizations

Popis výsledku v původním jazyce

CANDECOMP/PARAFAC (CP) has found numerous applications in wide variety of areas such as in chemometrics, telecommunication, data mining, neuroscience, separated representations. For an order- tensor, most CP algorithms can be computationally demanding due to computation of gradients which are related to products between tensor unfoldings and Khatri-Rao products of all factor matrices except one. These products have the largest workload in most CP algorithms. In this paper, we propose a fast method to deal with this issue. Themethod also reduces the extra memory requirements of CP algorithms. As a result, we can accelerate the standard alternating CP algorithms 20?30 times for order-5 and order-6 tensors, and even higher ratios can be obtained for higher order tensors (e.g., N>=10). The proposed method is more efficient than the state-of-the-art ALS algorithm which operates two modes at a time (ALSo2) in the Eigenvector PLS toolbox, especially for tensors with order N>=5 and high rank.

Název v anglickém jazyce

Fast Alternating LS Algorithms for High Order CANDECOMP/PARAFAC Tensor Factorizations

Popis výsledku anglicky

CANDECOMP/PARAFAC (CP) has found numerous applications in wide variety of areas such as in chemometrics, telecommunication, data mining, neuroscience, separated representations. For an order- tensor, most CP algorithms can be computationally demanding due to computation of gradients which are related to products between tensor unfoldings and Khatri-Rao products of all factor matrices except one. These products have the largest workload in most CP algorithms. In this paper, we propose a fast method to deal with this issue. Themethod also reduces the extra memory requirements of CP algorithms. As a result, we can accelerate the standard alternating CP algorithms 20?30 times for order-5 and order-6 tensors, and even higher ratios can be obtained for higher order tensors (e.g., N>=10). The proposed method is more efficient than the state-of-the-art ALS algorithm which operates two modes at a time (ALSo2) in the Eigenvector PLS toolbox, especially for tensors with order N>=5 and high rank.

Druh

J_x - Nezařazeno - Článek v odborném periodiku (Jimp, Jsc a Jost)

CEP obor

BB - Aplikovaná statistika, operační výzkum

OECD FORD obor

—

Projekt

<a href="/cs/project/GA102%2F09%2F1278" target="_blank" >GA102/09/1278: Pokročilé metody slepé separace signálu a slepé dekonvoluce</a><br>

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2013

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

IEEE Transactions on Signal Processing

ISSN

1053-587X

e-ISSN

—

Svazek periodika

61

Číslo periodika v rámci svazku

19

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

13

Strana od-do

4834-4846

Kód UT WoS článku

000324342900016

EID výsledku v databázi Scopus

—