Fast and efficient algorithms for nonnegative Tucker decomposition

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

19 Citations (Scopus)

Abstract

In this paper, we propose new and efficient algorithms for nonnegative Tucker decomposition (NTD): Fast α-NTD algorithm which is much precise and faster than α-NTD [1]; and β-NTD algorithm based on the β divergence. These new algorithms include efficient normalization and initialization steps which help to reduce considerably the running time and increase dramatically the performance. Moreover, the multilevel NTD scheme is also presented, allowing further improvements (almost perfect reconstruction). The performance was also compared to other well-known algorithms (HONMF, HOOI, ALS algorithms) for synthetic and real-world data as well.

Original languageEnglish
Title of host publicationAdvances in Neural Networks - ISNN 2008 - 5th International Symposium on Neural Networks, ISNN 2008, Proceedings
Pages772-782
Number of pages11
EditionPART 2
DOIs
Publication statusPublished - 2008
Externally publishedYes
Event5th International Symposium on Neural Networks, ISNN 2008 - Beijing, China
Duration: 24 Sep 200828 Sep 2008

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 2
Volume5264 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference5th International Symposium on Neural Networks, ISNN 2008
Country/TerritoryChina
CityBeijing
Period24/09/0828/09/08

Keywords

  • Alpha divergence
  • Beta divergence
  • Hierarchical decomposition
  • Nonnegative matrix factorization (NMF)
  • Nonnegative Tucker decomposition (NTD)

Fingerprint

Dive into the research topics of 'Fast and efficient algorithms for nonnegative Tucker decomposition'. Together they form a unique fingerprint.

Cite this