Non-negative sub-tensor ensemble factorization (NsTEF) algorithm. A new incremental tensor factorization for large data sets

Autor(es)

Vigneron, Vincent ^[1] ; Kodewitz, Andreas ^[1] ; da Costa, Michele Nazareth ^[2] ; Tome, Ana Maria ^[3] ; Langlang, Elmar ^[4]

Afiliação do(s) autor(es)

^[1] Univ Paris Saclay, Univ Ewy, IBISC, F-91025 Evry - France

^[2] Univ Campinas UNICAMP, DSPCom Lab, POB 6101, BR-13083852 Campinas, SP - Brazil

^[3] Univ Aveiro, Dept Elect Telecomunicacoes & Informat, Aveiro - Portugal

^[4] Univ Regensburg, Inst Biophys & Phys Biochem, Univ Str 31, D-93040 Regensburg - Germany

• Tipo de documento: Artigo Científico
• Fonte: Signal Processing; v. 144, p. 77-86, MAR 2018.
• Citações Web of Science: 1

Resumo

In this work we present a novel algorithm for nonnegative tensor factorization (NTF). Standard NTF algorithms are very restricted in the size of tensors that can be decomposed. Our algorithm overcomes this size restriction by interpreting the tensor as a set of sub-tensors and by proceeding the decomposition of sub-tensor by sub-tensor. This approach requires only one sub-tensor at once to be available in memory. (C) 2017 Elsevier B.V. All rights reserved. (AU)

• Processo FAPESP: 14/23936-4 - Aplicações de processamento de dados multidimensionais usando métodos tensoriais
• Beneficiário: Michele Nazareth da Costa
• Modalidade de apoio: Bolsas no Brasil - Pós-Doutorado