Improving the speed of multiway algorithms part II: Compression

Rasmus Bro; Claus A. Andersson

doi:10.1016/S0169-7439(98)00011-2

Improving the speed of multiway algorithms part II: Compression

Rasmus Bro^*, Claus A. Andersson

^*Corresponding author for this work

Food Analytics and Biotechnology

68 Citations (Scopus)

Abstract

In this paper an approach is developed for compressing a multiway array prior to estimating a multilinear model with the purpose of speeding up the estimation. A method is developed which seems very well-suited for a rich variety of models with optional constraints on the factors. It is based on three key aspects: (1) a fast implementation of a Tucker3 algorithm, which serves as the compression method, (2) the optimality theorem of the CANDELINC model, which ensures that the compressed array preserves the original variation maximally, and (3) a set of guidelines for how to incorporate optional constraints. The compression approach is tested on two large data sets and shown to speed up the estimation of the model up to 40 times. The developed algorithms can be downloaded from http:\\newton.mli.kvl.dk\foodtech.html.

Original language	English
Journal	Chemometrics and Intelligent Laboratory Systems
Volume	42
Issue number	1-2
Pages (from-to)	105-113
Number of pages	9
ISSN	0169-7439
DOIs	https://doi.org/10.1016/S0169-7439(98)00011-2
Publication status	Published - 24 Aug 1998

Keywords

CANDELINC
Constraints
Data compression
PARAFAC
Tucker1
Tucker3

Access to Document

10.1016/S0169-7439(98)00011-2

Cite this

@article{02092d249b30402a8c305ca074487d58,

title = "Improving the speed of multiway algorithms part II: Compression",

abstract = "In this paper an approach is developed for compressing a multiway array prior to estimating a multilinear model with the purpose of speeding up the estimation. A method is developed which seems very well-suited for a rich variety of models with optional constraints on the factors. It is based on three key aspects: (1) a fast implementation of a Tucker3 algorithm, which serves as the compression method, (2) the optimality theorem of the CANDELINC model, which ensures that the compressed array preserves the original variation maximally, and (3) a set of guidelines for how to incorporate optional constraints. The compression approach is tested on two large data sets and shown to speed up the estimation of the model up to 40 times. The developed algorithms can be downloaded from http:\\newton.mli.kvl.dk\foodtech.html.",

keywords = "CANDELINC, Constraints, Data compression, PARAFAC, Tucker1, Tucker3",

author = "Rasmus Bro and Andersson, {Claus A.}",

year = "1998",

month = aug,

day = "24",

doi = "10.1016/S0169-7439(98)00011-2",

language = "English",

volume = "42",

pages = "105--113",

journal = "Chemometrics and Intelligent Laboratory Systems",

issn = "0169-7439",

publisher = "Elsevier",

number = "1-2",

}

TY - JOUR

T1 - Improving the speed of multiway algorithms part II

T2 - Compression

AU - Bro, Rasmus

AU - Andersson, Claus A.

PY - 1998/8/24

Y1 - 1998/8/24

N2 - In this paper an approach is developed for compressing a multiway array prior to estimating a multilinear model with the purpose of speeding up the estimation. A method is developed which seems very well-suited for a rich variety of models with optional constraints on the factors. It is based on three key aspects: (1) a fast implementation of a Tucker3 algorithm, which serves as the compression method, (2) the optimality theorem of the CANDELINC model, which ensures that the compressed array preserves the original variation maximally, and (3) a set of guidelines for how to incorporate optional constraints. The compression approach is tested on two large data sets and shown to speed up the estimation of the model up to 40 times. The developed algorithms can be downloaded from http:\\newton.mli.kvl.dk\foodtech.html.

AB - In this paper an approach is developed for compressing a multiway array prior to estimating a multilinear model with the purpose of speeding up the estimation. A method is developed which seems very well-suited for a rich variety of models with optional constraints on the factors. It is based on three key aspects: (1) a fast implementation of a Tucker3 algorithm, which serves as the compression method, (2) the optimality theorem of the CANDELINC model, which ensures that the compressed array preserves the original variation maximally, and (3) a set of guidelines for how to incorporate optional constraints. The compression approach is tested on two large data sets and shown to speed up the estimation of the model up to 40 times. The developed algorithms can be downloaded from http:\\newton.mli.kvl.dk\foodtech.html.

KW - CANDELINC

KW - Constraints

KW - Data compression

KW - PARAFAC

KW - Tucker1

KW - Tucker3

UR - http://www.scopus.com/inward/record.url?scp=0032563557&partnerID=8YFLogxK

U2 - 10.1016/S0169-7439(98)00011-2

DO - 10.1016/S0169-7439(98)00011-2

M3 - Journal article

AN - SCOPUS:0032563557

SN - 0169-7439

VL - 42

SP - 105

EP - 113

JO - Chemometrics and Intelligent Laboratory Systems

JF - Chemometrics and Intelligent Laboratory Systems

IS - 1-2

ER -

Improving the speed of multiway algorithms part II: Compression

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this