Tensor Networks for Latent Variable Analysis: Higher Order Canonical Polyadic Decomposition

Anh Huy Phan, Andrzej Cichocki, Ivan Oseledets, Giuseppe G. Calvi, Salman Ahmadi-Asl, Danilo P. Mandic

    Research output: Contribution to journalArticlepeer-review

    4 Citations (Scopus)

    Abstract

    The canonical polyadic decomposition (CPD) is a convenient and intuitive tool for tensor factorization; however, for higher order tensors, it often exhibits high computational cost and permutation of tensor entries, and these undesirable effects grow exponentially with the tensor order. Prior compression of tensor in-hand can reduce the computational cost of CPD, but this is only applicable when the rank R of the decomposition does not exceed the tensor dimensions. To resolve these issues, we present a novel method for CPD of higher order tensors, which rests upon a simple tensor network of representative inter-connected core tensors of orders not higher than 3. For rigor, we develop an exact conversion scheme from the core tensors to the factor matrices in CPD and an iterative algorithm of low complexity to estimate these factor matrices for the inexact case. Comprehensive simulations over a variety of scenarios support the proposed approach.

    Original languageEnglish
    Article number8812695
    Pages (from-to)2174-2188
    Number of pages15
    JournalIEEE Transactions on Neural Networks and Learning Systems
    Volume31
    Issue number6
    DOIs
    Publication statusPublished - Jun 2020

    Keywords

    • Blind source separation
    • canonical polyadic decomposition
    • exact conversion
    • harmonic retrieval
    • higher order tensor
    • tensor network
    • tensor train

    Fingerprint

    Dive into the research topics of 'Tensor Networks for Latent Variable Analysis: Higher Order Canonical Polyadic Decomposition'. Together they form a unique fingerprint.

    Cite this