Parallel algorithms for large scale constrained tensor decomposition

Athanasios P. Liavas, Nikolaos Sidiropoulos

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Scopus citations

Abstract

Most tensor decomposition algorithms were developed for in-memory computation on a single machine. There are a few recent exceptions that were designed for parallel and distributed computation, but these cannot easily incorporate practically important constraints, such as nonnegativity. A new constrained tensor factorization framework is proposed in this paper, building upon the Alternating Direction method of Multipliers (ADMoM). It is shown that this simplifies computations, bypassing the need to solve constrained optimization problems in each iteration, yielding algorithms that are naturally amenable to parallel implementation. The methodology is exemplified using nonnegativity as a baseline constraint, but the proposed framework can incorporate many other types of constraints. Numerical experiments are encouraging, indicating that ADMoM-based nonnegative tensor factorization (NTF) has high potential as an alternative to state-of-the-art approaches.

Original languageEnglish (US)
Title of host publication2015 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages2459-2463
Number of pages5
Volume2015-August
ISBN (Electronic)9781467369978
DOIs
StatePublished - Jan 1 2015
Event40th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015 - Brisbane, Australia
Duration: Apr 19 2014Apr 24 2014

Other

Other40th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015
CountryAustralia
CityBrisbane
Period4/19/144/24/14

Keywords

  • CANDECOMP
  • PARAFAC
  • Tensors
  • constrained optimization
  • nonnegative factorization
  • parallel algorithms

Fingerprint Dive into the research topics of 'Parallel algorithms for large scale constrained tensor decomposition'. Together they form a unique fingerprint.

Cite this